From danbri at w3.org Wed Sep 1 16:13:28 2004 From: danbri at w3.org (Dan Brickley) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] stuff to look at Message-ID: <20040901161328.GB17604@homer.w3.org> some urls that have come up in conversation at FOAF Galway: http://web.mit.edu/ryanlee/www/thesis/thesis.html Personal Data Protection in the Semantic Web Masters of Engineering Thesis http://www.w3.org/TR/rdf-dawg-uc/ RDF Data Access Use Cases and Requirements http://www.w3.org/TR/2004/WD-swbp-n-aryRelations-20040721/ Defining N-ary Relations on the Semantic Web: Use With Individuals SKOS: http://www.w3.org/2004/02/skos/core/ RDFAuthor's RDF querying tutorial, http://rdfweb.org/people/damian/2001/10/RDFAuthor/Tutorial/Tutorial2/ From ronwalf at volus.net Thu Sep 2 16:10:33 2004 From: ronwalf at volus.net (Ron Alford) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] Using OWL instead of foaf:membershipClass property Message-ID: <413745F9.6070403@volus.net> I haven't seen much response to my suggestions to make foaf owl-friendlier, so I'm putting this forward both as to help make foaf semantics more machine readable, and to motivate some clean ups to make foaf friendlier to owl reasoners. IMHO, foaf:membershipClass is kind of hacky. It requires non-trivial ontology-specific reasoning which can be replaced by standard owl semantics. Let's take the example on the webpage: ILRT staff Already it's using fairly complex owl-isms. Also, I do not think it means what the page thinks it means, but that's a seperate issues (all ILRT staff have a certain work homepage, but others may have that same work homepage and not be staff). To make it into owl-dl, we'll have to add one property - memberOf, and declare it to be the inverse of foaf:member. ILRT Staff Now we have that ILRTStaffGroup has all ILRTStaff as members, without having to define our own reasoning directives. -Ron Alford From julian_bond at voidstar.com Thu Sep 2 18:49:41 2004 From: julian_bond at voidstar.com (Julian Bond) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] FOAF-Galway Message-ID: I'd just like to thank everyone and anyone involved in organising FOAF-Galway. I had a great time with a great bunch of people and learnt a lot. Well done. -- Julian Bond Email&MSM: julian.bond at voidstar.com Webmaster: http://www.ecademy.com/ Personal WebLog: http://www.voidstar.com/ M: +44 (0)77 5907 2173 T: +44 (0)192 0412 433 From nick.kings at bt.com Fri Sep 3 08:51:59 2004 From: nick.kings at bt.com (nick.kings@bt.com) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] FOAF-Galway Message-ID: <21DA6754A9238B48B92F39637EF307FD0964EBB7@i2km41-ukdy.domain1.systemhost.net> I'd like to second that! Thanks!!! Nick _____ Nicholas J. Kings (Nick), Next Generation Web Research, BT Exact Technologies, http://www.btexact.com/ _____ > -----Original Message----- > From: rdfweb-dev-bounces@vapours.rdfweb.org > [mailto:rdfweb-dev-bounces@vapours.rdfweb.org] On Behalf Of > Julian Bond > Sent: 02 September 2004 19:50 > To: rdfweb-dev@vapours.rdfweb.org > Subject: [rdfweb-dev] FOAF-Galway > > I'd just like to thank everyone and anyone involved in > organising FOAF-Galway. I had a great time with a great bunch > of people and learnt a lot. Well done. > > -- > Julian Bond Email&MSM: julian.bond at voidstar.com > Webmaster: http://www.ecademy.com/ > Personal WebLog: http://www.voidstar.com/ > M: +44 (0)77 5907 2173 T: +44 (0)192 0412 433 > > _______________________________________________ > rdfweb-dev mailing list > rdfweb-dev@vapours.rdfweb.org > wiki: http://rdfweb.org/topic/FoafProject > http://rdfweb.org/mailman/listinfo/rdfweb-dev > From ldodds at ingenta.com Fri Sep 3 11:25:17 2004 From: ldodds at ingenta.com (Leigh Dodds) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] [Fwd: FOAF IrcMeeting] Message-ID: <4138549D.80208@ingenta.com> resending as I was unsubscribed... -------- Original Message -------- Subject: FOAF IrcMeeting Date: Fri, 03 Sep 2004 10:40:57 +0100 From: Leigh Dodds To: rdfweb Comments re: regular FOAF IRC meetings. See also http://rdfweb.org/topic/IrcMeeting. To make the meetings most effective, can we agree on some basic procedures for carrying them out? - topics are published ~week in advance, ideally more so that comments from all communities can be included. - each chat is "lead" by a moderator, who is responsible for taking notes, and summarising the discussion. Perhaps the sponsor for a particular topic becomes the moderator. - each chat is summarised in a page in the Wiki, complete with list of actions/decisions - chat summary is circulated to this list and also linked from FOAF blog to ensure that people who couldn't attend (or don't use IRC) can be included. Common sense stuff I suppose, but doesn't hurt to write it down. L. From ldodds at ingenta.com Fri Sep 3 11:25:31 2004 From: ldodds at ingenta.com (Leigh Dodds) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] [Fwd: Process, Issue Tracker] Message-ID: <413854AB.7070205@ingenta.com> resending... -------- Original Message -------- Subject: Process, Issue Tracker Date: Fri, 03 Sep 2004 10:50:22 +0100 From: Leigh Dodds To: rdfweb Reading through the FOAFCommunityProcess page on the Wiki [1] I agreed whole-heartedly with the point that notes that having a clear decision trail/record is required. I'm just wondering whether we can make some improvements to the IssueTracker page to make things easier to find (and fix). One improvement that I'd personally find useful is breaking down the "issues" into categories: spec revision proposals (e.g. new terms), spec maintenance issues (e.g. docs), new areas to explore (e.g. PGP vocab), etc. Some of these are things that we could collaborativel resolve, i.e. anyone could propose draft text to better document a term, or provide patches to the schema to ensure human and machine-readable versions are in accord. It's the other issues that are the ones that require an audit history: why did we add/remove a term, why was it changed, etc. And it's these issues that require the most discussion. Would help to be able to list them all separately. Maybe we just just move a bunch of items to a new page? (Although I did like the bugzilla interface as that allows searching for items by category) Cheers, L. [1]. http://rdfweb.org/topic/FOAFCommunityProcess From crschmidt at crschmidt.net Fri Sep 3 15:29:08 2004 From: crschmidt at crschmidt.net (Christopher Schmidt) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] [Fwd: FOAF IrcMeeting] Message-ID: <20040903152908.GC5460@peanut.crschmidt.net> A few things from discussion today on IRC, as well as some responses. 1. Discussions on the topic list at IrcMeeting right now are foaf related, but really apply more generally to RDF. It might be better to hold the meetings in #rdfig, in terms of attracting more interested people as well as having a chump, keeping everyone involved, etc. 2. These meetings can (and probably will) be completely seperate from the FoafCommunityProcess meetings, designed for extension of the ontology: [[[ * agreed that a regular 'heartbeat' for status/progress updates to FOAF spec (issue lists, TODOs etc) would be helpful. * agreed an initial IRC meeting 1700pm BST 2004-09-15 in #foaf, for status/update from editor; danbri to circulate a meeting agenda to rdfweb-dev. ]]] -- http://rdfweb.org/topic/FOAFCommunityProcess Although these meetings fall into a similar arena, there is no need to discuss updates to the schema on a weekly basis. However, with a wide range of topics, weekly meetings along the lines of "interaction between $topic and FOAF" don't seem out of the question. > Comments re: regular FOAF IRC meetings. See also > http://rdfweb.org/topic/IrcMeeting. > > To make the meetings most effective, can we agree on > some basic procedures for carrying them out? > > - topics are published ~week in advance, ideally more > so that comments from all communities can be included. > > - each chat is "lead" by a moderator, who is responsible > for taking notes, and summarising the discussion. Perhaps > the sponsor for a particular topic becomes the moderator. > > - each chat is summarised in a page in the Wiki, complete with > list of actions/decisions > > - chat summary is circulated to this list and also linked from > FOAF blog to ensure that people who couldn't attend (or don't > use IRC) can be included. > > Common sense stuff I suppose, but doesn't hurt to write it down. > > L. I have no problems with any of these suggestions, and since I'm one to open my big mouth a lot and not do a lot about it, I'd be glad to act as a moderator for at least the first meeting. Since I do want to get this started, and not to interfere with the ontology discussion on the 15th, I'd like to propose that we wait until Noon GMT, Sunday to see if there's a definite preference for a topic. If there isn't, I'll make an executive decision as a moderator of the talk as to which topic will be discussed, in an effort to allow people to gather whatever thoughts they want to share. The agenda I had in mind would be something along the lines of the following (times in Eastern because that's what I'm thinking in): 11:00AM Call to Order, inform other people meeting is starting Introductions: Name, Rank, Serial Number 11:05AM Topic introduction, linking to relevant resources to get everyone involved on the same page 11:15AM Discussion: What people would like to see, what's needed, what's missing, what's okay but needs more work, general topic chatter 11:40AM Wrap up - move important/relevant links to wiki, add participants list. Define goals - what do people want to see, what needs to be done, "Action Items", also added to wiki 11:50AM Discussion of next meeting topic Then an email could be sent to the list detailing what people 1. Are working on 2. Will be working on 3. Discussed 4. Think about the topic Obviously, these times are relatively fluid, if the discussion needs or wants to continue people can either break off and work on it or something similar. I figure an hour is the longest a group of people can reasonably devote at random times of the day, although if it needs to go longer it can. Does this seem like a reasonable goal and timeline? Anyone have any objections? Anyone want to participate? :) The time of the meeting in many locations: http://tinyurl.com/65f37 -- Christopher Schmidt From Libby.Miller at bristol.ac.uk Sat Sep 4 19:32:48 2004 From: Libby.Miller at bristol.ac.uk (Libby Miller) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] FOAF-Galway In-Reply-To: References: Message-ID: hey Julian! I'm glad you could make it and that you had a good time. It was great to see you :) Thanks to everyone who attended. I had a really productive and fun time, and I hope you did too (we'll give you the opportunity to comment on that soon I hope). Libby On Thu, 2 Sep 2004, Julian Bond wrote: > I'd just like to thank everyone and anyone involved in organising > FOAF-Galway. I had a great time with a great bunch of people and learnt a > lot. Well done. > > -- > Julian Bond Email&MSM: julian.bond at voidstar.com > Webmaster: http://www.ecademy.com/ > Personal WebLog: http://www.voidstar.com/ > M: +44 (0)77 5907 2173 T: +44 (0)192 0412 433 > > _______________________________________________ > rdfweb-dev mailing list > rdfweb-dev@vapours.rdfweb.org > wiki: http://rdfweb.org/topic/FoafProject > http://rdfweb.org/mailman/listinfo/rdfweb-dev > From Libby.Miller at bristol.ac.uk Sat Sep 4 19:40:16 2004 From: Libby.Miller at bristol.ac.uk (Libby Miller) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] FOAF-Galway In-Reply-To: <21DA6754A9238B48B92F39637EF307FD0964EBB7@i2km41-ukdy.domain1.systemhost.net> References: <21DA6754A9238B48B92F39637EF307FD0964EBB7@i2km41-ukdy.domain1.systemhost.net> Message-ID: Thanks for coming Nick :) Libby On Fri, 3 Sep 2004 nick.kings@bt.com wrote: > > I'd like to second that! > > Thanks!!! > > Nick > > > > _____ > > Nicholas J. Kings (Nick), > Next Generation Web Research, BT Exact Technologies, > http://www.btexact.com/ > > _____ > > > >> -----Original Message----- >> From: rdfweb-dev-bounces@vapours.rdfweb.org >> [mailto:rdfweb-dev-bounces@vapours.rdfweb.org] On Behalf Of >> Julian Bond >> Sent: 02 September 2004 19:50 >> To: rdfweb-dev@vapours.rdfweb.org >> Subject: [rdfweb-dev] FOAF-Galway >> >> I'd just like to thank everyone and anyone involved in >> organising FOAF-Galway. I had a great time with a great bunch >> of people and learnt a lot. Well done. >> >> -- >> Julian Bond Email&MSM: julian.bond at voidstar.com >> Webmaster: http://www.ecademy.com/ >> Personal WebLog: http://www.voidstar.com/ >> M: +44 (0)77 5907 2173 T: +44 (0)192 0412 433 >> >> _______________________________________________ >> rdfweb-dev mailing list >> rdfweb-dev@vapours.rdfweb.org >> wiki: http://rdfweb.org/topic/FoafProject >> http://rdfweb.org/mailman/listinfo/rdfweb-dev >> > > _______________________________________________ > rdfweb-dev mailing list > rdfweb-dev@vapours.rdfweb.org > wiki: http://rdfweb.org/topic/FoafProject > http://rdfweb.org/mailman/listinfo/rdfweb-dev > From Libby.Miller at bristol.ac.uk Sun Sep 5 00:07:51 2004 From: Libby.Miller at bristol.ac.uk (Libby Miller) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] more photos from foaf-galway Message-ID: http://swordfish.rdfweb.org/photos/2004/08/31/ http://swordfish.rdfweb.org/photos/2004/09/01/ http://swordfish.rdfweb.org/photos/2004/09/02/ http://swordfish.rdfweb.org/photos/2004/09/03/ Libby From danny.ayers at gmail.com Sun Sep 5 19:08:00 2004 From: danny.ayers at gmail.com (Danny Ayers) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] Re: FoafInBrief (info towards Atom PersonConstruct) In-Reply-To: References: <1f2ed5cd04090509585ba3105b@mail.gmail.com> Message-ID: <1f2ed5cd0409051208b547e52@mail.gmail.com> On Sun, 05 Sep 2004 11:16:01 -0700, Tim Bray wrote: > On Sep 5, 2004, at 9:58 AM, Danny Ayers wrote: > In my experience FOAF is well thought-out and (speaking personally) I'd > be happy to avoid redesigning personal info in Atom by getting help > from FOAF. But author-identification is so basic to Atom that dipping > into another namespace seems egregious, so we'd want to use FOAF by > copy rather than by reference. Yep, that sounds reasonable. The path of least resistance being to learn what we can from FOAF's identification system, and implement something specific to Atom based on those lessons. My hope is that a side effect will be a fairly simple way of working with Atom and FOAF within the same application. Dan will hopefully fill in the details (he was flying to Tokyo today, so it probably won't be immediate), but an overhaul of the vocabulary and documentation was scheduled to follow the summer events - FOAFCamp in the Netherlands and a FOAF workshop in Galway. The latter only finished a couple of days go, but from what I gather a roadmap was on the agenda, and increased standardisation was under discussion. Cheers, Danny. -- http://dannyayers.com From danbri at w3.org Mon Sep 6 06:12:43 2004 From: danbri at w3.org (Dan Brickley) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] Re: FoafInBrief (info towards Atom PersonConstruct) In-Reply-To: <1f2ed5cd0409051208b547e52@mail.gmail.com> References: <1f2ed5cd04090509585ba3105b@mail.gmail.com> <1f2ed5cd0409051208b547e52@mail.gmail.com> Message-ID: <20040906061243.GB7484@homer.w3.org> * Danny Ayers [2004-09-05 21:08+0200] > On Sun, 05 Sep 2004 11:16:01 -0700, Tim Bray wrote: > > On Sep 5, 2004, at 9:58 AM, Danny Ayers wrote: > > > In my experience FOAF is well thought-out and (speaking personally) I'd > > be happy to avoid redesigning personal info in Atom by getting help > > from FOAF. But author-identification is so basic to Atom that dipping > > into another namespace seems egregious, so we'd want to use FOAF by > > copy rather than by reference. > > Yep, that sounds reasonable. The path of least resistance being to > learn what we can from FOAF's identification system, and implement > something specific to Atom based on those lessons. My hope is that a > side effect will be a fairly simple way of working with Atom and FOAF > within the same application. > > Dan will hopefully fill in the details (he was flying to Tokyo today, > so it probably won't be immediate), but an overhaul of the vocabulary > and documentation was scheduled to follow the summer events - FOAFCamp > in the Netherlands and a FOAF workshop in Galway. The latter only > finished a couple of days go, but from what I gather a roadmap was on > the agenda, and increased standardisation was under discussion. Hi. Made it to Tokyo, though my brain isn't all here yet. I would be pleased to see a defined way of getting from Atom's notion of person to FOAF's, and back again. The core idea in FOAF for IDs is to use any properties that are uniquely identifying, rather than having silly angels-on-pinheads arguments about "what the URI is for a person". We say that 'homepage', 'weblog', (personal)'mbox' and 'mbox_sha1sum' are such things. There are also a few other pieces of FOAF that can play interestingly in an RSS1 environment, such as 'topic' (a relation between a document and something it is about); 'primaryTopic' (a relation between a topic and the main thing it is about), 'img' a relation between [from memory] a person and an image that is particularly characterstic of it. Also 'depicts' is popular for image description apps (moblogging etc), and relates an Image to a thing it depicts. These combine in ways that go a bit beyond people-description, eg. you can say that a Document has as its primaryTopic a thing that has a homepage of http://www.somemovie.example.com/. The idea is to emphasise idioms that allow for data merging across very loosly coordinated systems, by piggybacking off of very well known identifiers. Other person description things include workplaceHomepage, which is a basic construct that should allow cluetrainish apps such as 'find me all the things that are of type rss:channel and which are made by persons whose workplaceHomepage is http://www.sun.com/ or http://www.microsoft.com/. I doubt Atom'd want to go into such detail in its core, but as we get a process together (Galway notes are in http://rdfweb.org/topic/FOAFCommunityProcess), we will start tagging bits of FOAF as 'stable' and that might give groups such as Atom a bit more evidence that they're reasonable things to cite or copy. I think there's a good case for publishing something through W3C (eg. as an Interest Group Note of the SW IG) and/or a snapshot of term definitions in IETF somehow. I'm currently pretty lukewarm on the idea of transferring the whole thing to W3C, but have been discussing the idea of making a task force of the Semantic Web Best Practices WG that uses FOAF as a case study of how to (or not to!) make SW vocabs, and the issues that arise. Back re Atom, perhaps GRDDL, http://www.w3.org/TR/2004/NOTE-grddl-20040413/ would be a way of characterising a mapping from Atom's syntax into something RDF tools could eat... I do worry about people getting confused by seeing bits of FOAF RDF appear in non-RDF syntax, but if the FOAF subtrees of the Atom file used RDF notation, maybe that'd work. I'll have a think. Certainly for basic stuff like 'Person','mbox','mbox_sha1sum','homepage', 'weblog' there's a decent case for copy rather than reference. I need to have another look at latest Atom/person stuff before I can be any more useful here. It might not get gotten to until I'm back in Bristol on 14/15th... hope this helps, cheers, Dan ps. btw I'm not sure we say in the spec (we should) but FOAF owes a great deal to MCF, especially Guha and TimBray's MCF-in-XML spec... From henry.story at bblfish.net Mon Sep 6 15:23:26 2004 From: henry.story at bblfish.net (Henry Story) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] Re: FoafInBrief (info towards Atom PersonConstruct) In-Reply-To: <20040906061243.GB7484@homer.w3.org> References: <1f2ed5cd04090509585ba3105b@mail.gmail.com> <1f2ed5cd0409051208b547e52@mail.gmail.com> <20040906061243.GB7484@homer.w3.org> Message-ID: First thanks a lot to Danny Ayers for the hard work he put into his clear and simple explanation of RDF and FOAF [1]. That is a very nice introduction to the technology. The best way to learn about it is to play with it a little. Making one's own FOAF file is a good first step. To those embarking down that road I seriously recommend writing the code in N3 and using a tool like cwm [2] to convert back and forth from that notation to the xml format. Take it from a newbie like me: it is not that difficult, and it's a lot of fun. On the topic at hand, I think there is a good beginning of a consensus being arrived at here: - the FOAF group is doing a lot of good work on the Person concept, - some core elements of that concept are required in Atom but importing all of the FOAF concepts might be overkill -> we just need to make sure we have a good interface between the Atom Person Construct and the FOAF equivalent (foaf:Agent or foaf:Person?) The above argument would tend to go against my model of Atom+OWL [3] where I directly import the FOAF ontology, and favor Danny's model [4] where he just defines the Person element inside of the atom namespace. In Java OO programming terms we can think of this as Danny creating a Person interface in the atom.* package, which all the other atom classes refer to when they want to speak of a Person object. What we need is a simple implementation class somewhere that implements that Person interface with the foaf equivalent. I would be really interested to know what the best way to do this in RDF is. One way to do this would be to have Atom:Person be a subclass of foaf:Person (or foaf:Agent?). Would there be better methods for this? I have set up a google newsgroup [5] for discussion of those issues that may appear to be more tangential to this group. Henry Story [1] http://www.intertwingly.net/wiki/pie/FoafInBrief [2] http://www.w3.org/2000/10/swap/doc/cwm.html [3] http://bblfish.net/work/atom-owl/2004-08-12/blogexample.html [4] http://semtext.org/atom/atom.html (section 3.2) [5] http://groups-beta.google.com/group/atom-owl From GK at ninebynine.org Mon Sep 6 15:22:15 2004 From: GK at ninebynine.org (Graham Klyne) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] Postal address vocabulary In-Reply-To: <413323AF.6080503@gmuer.ch> Message-ID: <5.1.0.14.2.20040906153628.01fee8c8@127.0.0.1> At 14:55 30/08/04 +0200, Reto Bachmann-Gmuer wrote: >I've just added a proposal for a vocabulary to add postal addresses to >foaf-profiles at http://rdfweb.org/topic/AddressVocab > >Any comment would be appreciated! This appears to be a comprehensive approach, and I wonder if too comprehensive for a lightweight, general-use vocabulary like FOAF. ... For what it's worth, I informally proposed some FOAF terms for postal addresses some time ago, in support of some work I have been doing to use RDF in the generation of data for an IETF protocol element registry. An example of the terms I have used is: [[ hdr:Author_PR a wn:Person ; foaf:name "Peter W. Resnick" ; foaf:mbox ; foaf:organization "QUALCOMM Incorporated" ; foaf:workplacePostal [ foaf:street1 "5775 Morehouse Drive" ; foaf:city "San Diego" ; foaf:postcode "CA 92121-1714" ; foaf:country "USA" ] ; foaf:workplaceTel "+1 858 651 4478" ; foaf:workplaceFax "+1 858 651 1102" . ]] A more complete indication of the vocabulary might be gleaned from this query expression which is part of my software implementation: [[ @query hrep:HdrPersonPattern ( ?person ( foaf:name ?aname & ( foaf:mbox ?ambox | foaf:homepage ?ahomepage ) & [ foaf:organization ?orgname ] & [ @hrep:HdrPostalPattern ] & [ foaf:workplaceTel ?wtel ] & [ foaf:workplaceFax ?wfax ] & [ foaf:workplaceHomepage ?wurl ] ) ) @query hrep:HdrPostalPattern ( foaf:workplacePostal ?wp ( [ foaf:building ?wpbuilding ] & [ foaf:street1 ?wpstreet1 ] & [ foaf:street2 ?wpstreet2 ] & [ foaf:city ?wpcity ] & [ foaf:area ?wparea ] & [ foaf:postcode ?wppostcode ] & [ foaf:country ?wpcountry ] ) ) ]] I'm not suggesting this is any better than your proposal. I'm just mentioning it here because it was derived from a specific use-case (capturing details of IETF document editors), and as such might be a useful touchstone. #g ------------ Graham Klyne For email: http://www.ninebynine.org/#Contact From danny.ayers at gmail.com Mon Sep 6 17:07:31 2004 From: danny.ayers at gmail.com (Danny Ayers) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] Re: FoafInBrief (info towards Atom PersonConstruct) In-Reply-To: References: <1f2ed5cd04090509585ba3105b@mail.gmail.com> <1f2ed5cd0409051208b547e52@mail.gmail.com> <20040906061243.GB7484@homer.w3.org> Message-ID: <1f2ed5cd040906100720259d6@mail.gmail.com> On Mon, 6 Sep 2004 17:23:26 +0200, Henry Story wrote: Thanks Henry, just a quick comment - > One way to do this would be to have Atom:Person be a subclass of > foaf:Person (or foaf:Agent?). Would there be better methods for this? I > have set up a google newsgroup [5] for discussion of those issues that > may appear to be more tangential to this group. Before anyone starts screaming, I would assume that the mapping will only be used in the RDF world, no dependencies would be introduced into Atom. Alternatives that spring to mind are stating that foaf:Person (/foaf:Agent) is an rdfs:subClassOf of atom:Person (the PersonConstruct), or each is an rdfs:subClassOf of the other (i.e. owl:equivalentClass). The current description of atom:Person is probably closer to foaf:Agent in that it may be an organisation. foaf:Agent is a little wider, it could also be something like a bot - maybe this broadening might be considered for the Atom construct. Cheers, Danny. -- http://dannyayers.com From marc at canter.com Tue Sep 7 14:27:33 2004 From: marc at canter.com (Marc Canter) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] Re: FoafInBrief (info towards Atom PersonConstruct) In-Reply-To: Message-ID: <004401c494e6$d67b24c0$650a0a0a@Lucy> Connecting FOAF and Atom together is crucial. For the record - we've attempted to create a simplified early stage implementation of FOAF - called the FOAFnet.org. http://www.socialtext.net/foafnet/index.cgi In that spec - we start with a very simple iteration of FOAF, including ONLY: - name - face - email (sha1sum) - website - list of friends The goal of FOAFnet was to facilitate exchange of entire social networks. Ecademy demonstrated it working at FOAF Galway and Tribe.net will ship with it - in a couple of weeks. Whatever the Atom folks come up with - we can support in FOAF. Let's work towards smooth transparent interop - to make our end-user's lives better. - Marc Canter -----Original Message----- From: rdfweb-dev-bounces@vapours.rdfweb.org [mailto:rdfweb-dev-bounces@vapours.rdfweb.org] On Behalf Of Henry Story Sent: Monday, September 06, 2004 5:23 PM To: Atom Syntax Cc: foafnet@yahoogroups.com; rdfweb-dev@vapours.rdfweb.org; Tim Bray; Paul Hoffman / IMC Subject: [rdfweb-dev] Re: FoafInBrief (info towards Atom PersonConstruct) First thanks a lot to Danny Ayers for the hard work he put into his clear and simple explanation of RDF and FOAF [1]. That is a very nice introduction to the technology. The best way to learn about it is to play with it a little. Making one's own FOAF file is a good first step. To those embarking down that road I seriously recommend writing the code in N3 and using a tool like cwm [2] to convert back and forth from that notation to the xml format. Take it from a newbie like me: it is not that difficult, and it's a lot of fun. On the topic at hand, I think there is a good beginning of a consensus being arrived at here: - the FOAF group is doing a lot of good work on the Person concept, - some core elements of that concept are required in Atom but importing all of the FOAF concepts might be overkill -> we just need to make sure we have a good interface between the Atom Person Construct and the FOAF equivalent (foaf:Agent or foaf:Person?) The above argument would tend to go against my model of Atom+OWL [3] where I directly import the FOAF ontology, and favor Danny's model [4] where he just defines the Person element inside of the atom namespace. In Java OO programming terms we can think of this as Danny creating a Person interface in the atom.* package, which all the other atom classes refer to when they want to speak of a Person object. What we need is a simple implementation class somewhere that implements that Person interface with the foaf equivalent. I would be really interested to know what the best way to do this in RDF is. One way to do this would be to have Atom:Person be a subclass of foaf:Person (or foaf:Agent?). Would there be better methods for this? I have set up a google newsgroup [5] for discussion of those issues that may appear to be more tangential to this group. Henry Story [1] http://www.intertwingly.net/wiki/pie/FoafInBrief [2] http://www.w3.org/2000/10/swap/doc/cwm.html [3] http://bblfish.net/work/atom-owl/2004-08-12/blogexample.html [4] http://semtext.org/atom/atom.html (section 3.2) [5] http://groups-beta.google.com/group/atom-owl _______________________________________________ rdfweb-dev mailing list rdfweb-dev@vapours.rdfweb.org wiki: http://rdfweb.org/topic/FoafProject http://rdfweb.org/mailman/listinfo/rdfweb-dev From mof-rdf at mfd-consult.dk Tue Sep 7 19:51:53 2004 From: mof-rdf at mfd-consult.dk (Morten Frederiksen) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] more photos from foaf-galway In-Reply-To: References: Message-ID: <200409072151.53369.mof-rdf@mfd-consult.dk> Hi all, Just a quick note of thanks to all at FOAF Galway, but of course especially to the organisers, and a pointer to a few more photos -- annotations are in the queue... http://www.wasab.dk/morten/2004/08/photos/eire/ Regards, Morten From gk at ninebynine.org Wed Sep 8 15:42:08 2004 From: gk at ninebynine.org (Graham Klyne) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] Haskell (FOAFcamp follow-up) Message-ID: <5.1.0.14.2.20040908160809.02a41ea0@127.0.0.1> Following my brief show'n'tell at the recent FOAFcamp of a "literate Haskell" work-in-progress for exploring description logic reasoning [1], Tony Bowden has passed me a reference to a description of an experiment in which the US military surveyed a range of programming lanugages, and didn't believe that the Haskell solution was real: It is significant that Mr. Domanski, Mr. Banowetz and Dr. Brosgol were all surprised and suspicious when we told them that Haskell prototype P1 (see appendix B) is a complete tested executable program. We provided them with a copy of P1 without explaining that it was a program, and based on preconceptions from their past experience, they had studied P1 under the assumption that it was a mixture of requirements specification and top level design. They were convinced it was incomplete because it did not address issues such as data structure design and execution order." "Haskell vs. Ada vs. C++ vs. Awk vs. ..., An Experiment in Software Prototyping Productivity" http://www.haskell.org/papers/NSWC/jfp.ps If/when XML and RDF library support for Haskell is stable and easily available, I'd hope for some similar advantages from Haskell for semantic web application development. #g -- [1] http://www.ninebynine.org/Software/HaskellDL/DLExploration.lhs ------------ Graham Klyne For email: http://www.ninebynine.org/#Contact From danny.ayers at gmail.com Mon Sep 13 11:26:31 2004 From: danny.ayers at gmail.com (Danny Ayers) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] RDF model sync, HTTP deltas, syndicated feeds (and a bounty) Message-ID: <1f2ed5cd0409130426540c86f9@mail.gmail.com> Long story short, PubSub.com is willing to offer a $5,000 bounty to whoever builds the "best" implementation of RFC3229 (Delta encoding in HTTP) with appropriate extensions for use with syndication feeds as an open source Apache module by Jan 1, 2005. The requirements for feed-oriented delta transfers are very, very close to those for passing around potentially huge, dynamically changing RDF/XML files, effectively synchronizing models between systems. Seems to me PubSub's solution could feed two birds with one bean, hence this mail. In the past few days there have been a couple of related developments - MSDN got stung with excessive bandwidth costs trying to broadcast the aggregated content of 1300+ feeds [2], and a (relatively!) simple algorithm for a delta-based solution has been described [3]. Most of the discussion has been taking place on the atom-syntax list [4]. Bob Wyman (of PubSub) describes the feed-oriented requirements some more at [5], there's a little more background on my blog at [6],[7]. Cheers, Danny. [1] http://www.ietf.org/rfc/rfc3229.txt [2] http://radio.weblogs.com/0001011/2004/09/08.html#a8195 [3] http://www.intertwingly.net/blog/2004/09/11/Vary-ETag [4] http://www.imc.org/atom-syntax/ [5] http://bobwyman.pubsub.com/main/2004/09/using_rfc3229_w.html [6] http://dannyayers.com/archives/2004/09/12/passing-things-around/ [7] http://dannyayers.com/archives/2004/09/13/geronimo/ -- http://dannyayers.com From reto at gmuer.ch Mon Sep 13 15:32:39 2004 From: reto at gmuer.ch (Reto Bachmann-Gmuer) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] Postal address vocabulary In-Reply-To: <5.1.0.14.2.20040906153628.01fee8c8@127.0.0.1> References: <5.1.0.14.2.20040906153628.01fee8c8@127.0.0.1> Message-ID: <4145BD97.7030501@gmuer.ch> Hi Graham, >> foaf-profiles at http://rdfweb.org/topic/AddressVocab > This appears to be a comprehensive approach, and I wonder if too > comprehensive for a lightweight, general-use vocabulary like FOAF. It's true that it long, I think a subset of the properties could be defined wich would be enough for most addresses. Maybe complexity is not foaf-style but I like the following aspects which I think are foafish: - decentral: separating post-operator information from the "objective" location makes it possible to be used independently of controlled systems. - culturale independence (this comes from copying from the S42 standard): any address can be represented and mechanism for locale specific rendering exist. - suitable for inference: it is clear if two persons live in the same house or street which may be hard if this information is joined in one literal. Of course aspects like sorting by street may be quite irrelevant in the foaf-world, but I think if complexity remains somehow reasonable it makes good sense to have one vocabulary for addresses which can be used in different areas (foaf, data-mining, post, geo) rather than multiple address-schemas. > > ... > > For what it's worth, I informally proposed some FOAF terms for postal > addresses some time ago, in support of some work I have been doing to > use RDF in the generation of data for an IETF protocol element > registry. An example of the terms I have used is: > [[ > hdr:Author_PR a wn:Person ; > foaf:name "Peter W. Resnick" ; > foaf:mbox ; > foaf:organization "QUALCOMM Incorporated" ; > foaf:workplacePostal > [ foaf:street1 "5775 Morehouse Drive" ; > foaf:city "San Diego" ; > foaf:postcode "CA 92121-1714" ; > foaf:country "USA" ] ; > foaf:workplaceTel "+1 858 651 4478" ; > foaf:workplaceFax "+1 858 651 1102" . > ]] I see that you include both organization and individual, which makes sense. I think that foaf should offer a vocabulary to express this, the address vocabulary I propopose for this at http://rdfweb.org/topic/RoleVocab. The idea is that an Agent participates in a Group having a certain Role in this group and that this Participation is itself a foaf:Agent. Like this the Participation could could be the adresse of an Adress, and of course have things like foaf:phone, foaf:knows which may or may not be the same as those of the individual or those of the group. I think it would be a pity if foaf would model an old-fashioned one person - one workplace world. cheers, reto From gk at ninebynine.org Mon Sep 13 19:08:56 2004 From: gk at ninebynine.org (Graham Klyne) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] Postal address vocabulary In-Reply-To: <4145BD97.7030501@gmuer.ch> References: <5.1.0.14.2.20040906153628.01fee8c8@127.0.0.1> <5.1.0.14.2.20040906153628.01fee8c8@127.0.0.1> Message-ID: <5.1.0.14.2.20040913200604.02b9aee0@127.0.0.1> Reto, I acknowledge all the points you make in your response. My question (and this is a question, not an assertion rhetorically styled as a question) is this: should this kind of detailed address-description vocabulary be part of FOAF, or a separate vocabulary (e.g. another namespace, which could, of course, be used *with* FOAF)? #g -- At 17:32 13/09/04 +0200, Reto Bachmann-Gmuer wrote: >Hi Graham, > >>>foaf-profiles at http://rdfweb.org/topic/AddressVocab > >>This appears to be a comprehensive approach, and I wonder if too >>comprehensive for a lightweight, general-use vocabulary like FOAF. >It's true that it long, I think a subset of the properties could be >defined wich would be enough for most addresses. > >Maybe complexity is not foaf-style but I like the following aspects >which I think are foafish: >- decentral: separating post-operator information from the "objective" >location makes it possible to be used independently of controlled systems. >- culturale independence (this comes from copying from the S42 >standard): any address can be represented and mechanism for locale >specific rendering exist. >- suitable for inference: it is clear if two persons live in the same >house or street which may be hard if this information is joined in one >literal. > >Of course aspects like sorting by street may be quite irrelevant in the >foaf-world, but I think if complexity remains somehow reasonable it >makes good sense to have one vocabulary for addresses which can be used >in different areas (foaf, data-mining, post, geo) rather than multiple >address-schemas. > >>... >>For what it's worth, I informally proposed some FOAF terms for postal >>addresses some time ago, in support of some work I have been doing to use >>RDF in the generation of data for an IETF protocol element registry. An >>example of the terms I have used is: >>[[ >>hdr:Author_PR a wn:Person ; >> foaf:name "Peter W. Resnick" ; >> foaf:mbox ; >> foaf:organization "QUALCOMM Incorporated" ; >> foaf:workplacePostal >> [ foaf:street1 "5775 Morehouse Drive" ; >> foaf:city "San Diego" ; >> foaf:postcode "CA 92121-1714" ; >> foaf:country "USA" ] ; >> foaf:workplaceTel "+1 858 651 4478" ; >> foaf:workplaceFax "+1 858 651 1102" . >>]] >I see that you include both organization and individual, which makes >sense. I think that foaf should offer a vocabulary to express this, the >address vocabulary I propopose for this at >http://rdfweb.org/topic/RoleVocab. The idea is that an Agent participates >in a Group having a certain Role in this group and that this Participation >is itself a foaf:Agent. Like this the Participation could could be the >adresse of an Adress, and of course have things like foaf:phone, >foaf:knows which may or may not be the same as those of the individual or >those of the group. I think it would be a pity if foaf would model an >old-fashioned one person - one workplace world. > >cheers, >reto ------------ Graham Klyne For email: http://www.ninebynine.org/#Contact From stpeter at jabber.org Mon Sep 13 21:50:11 2004 From: stpeter at jabber.org (Peter Saint-Andre) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] Re: RDF model sync, HTTP deltas, syndicated feeds (and a bounty) References: <1f2ed5cd0409130426540c86f9@mail.gmail.com> Message-ID: In article <1f2ed5cd0409130426540c86f9@mail.gmail.com>, Danny Ayers wrote: > http://bobwyman.pubsub.com/main/2004/09/using_rfc3229_w.html Bob also mentions that a true pubsub approach would be best in the long term; for instance, see: http://www.ietf.org/internet-drafts/draft-saintandre-atompub-notify-01.tx t /psa From zednenem at psualum.com Tue Sep 14 02:38:11 2004 From: zednenem at psualum.com (David Menendez) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] Postal address vocabulary In-Reply-To: <5.1.0.14.2.20040913200604.02b9aee0@127.0.0.1> Message-ID: Graham Klyne writes: > I acknowledge all the points you make in your response. My question > (and this is a question, not an assertion rhetorically styled as a > question) is this: should this kind of detailed address-description > vocabulary be part of FOAF, or a separate vocabulary (e.g. another > namespace, which could, of course, be used *with* FOAF)? I'd make it a separate vocabulary (call it an adjunct or a satellite vocabulary). FOAF and the address vocab are both large enough that they'd be easier to manage separately, and making the address vocab separate might make it more attractive to other projects. The only negative point which comes to mind is that it's slightly more work for people generating FOAF profiles by hand. -- David Menendez | "In this house, we obey the laws | of thermodynamics!" From julian_bond at voidstar.com Tue Sep 14 07:27:21 2004 From: julian_bond at voidstar.com (Julian Bond) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] Postal address vocabulary In-Reply-To: <5.1.0.14.2.20040913200604.02b9aee0@127.0.0.1> References: <5.1.0.14.2.20040906153628.01fee8c8@127.0.0.1> <5.1.0.14.2.20040906153628.01fee8c8@127.0.0.1> <4145BD97.7030501@gmuer.ch> <5.1.0.14.2.20040913200604.02b9aee0@127.0.0.1> Message-ID: Graham Klyne wrote: >I acknowledge all the points you make in your response. My question >(and this is a question, not an assertion rhetorically styled as a >question) is this: >should this kind of detailed address-description vocabulary be part of >FOAF, or a separate vocabulary (e.g. another namespace, which could, of >course, be used *with* FOAF)? I consider a generalised RDF (and/or XML) Address schema to be so important that it's development *must* be done in the context of a major standards organization. Given the history, rfc status and wide deployment and implementation of vCard I find it pretty hard to argue for a new one. So maybe what's really needed is a new vCard RFC that supersedes RFC 2426. If the problem is with the RDF representation of vCard, then that's being done within the W3C. If it's wrong there are well established ways to improve it. So I'm not against new projects to replace and redefine the same areas as existing schemas. But for something with such wide use I'd want to see a lot of justification and at least some comment about how you expect to get traction for the new standard. There's already a lot of FOAF out there that contains data in the vcard schema as well as many other schemas. And this is a key benefit of RDF. So it really doesn't look like there's a need to keep adding to the FOAF schema tags that overlap with other existing schemas. I've argued before that actually FOAF should be slimmed down so that it becomes a structural schema with only just enough data tags to make the structure useful and readable. Taken to an extreme (ie too far!) that would mean foaf:Person, foaf:mbox_sha1sum, foaf:name, foaf:knows and that's about all. Now that's clearly a mistake, but I am coming round to the idea that we shouldn't confuse the FOAF schema as structure by adding more and more general purpose tags that happen to be missing from the existing schema space. Which is all a long winded way of saying, if you see a hole create a new namespace. And then fight for mindshare for that namespace in the ecology of namespace standards. David Menendez wrote: >The only negative point which comes to mind is that it's slightly more >work for people generating FOAF profiles by hand. The only people generating FOAF by hand should be experimenters and programmer/implementors with an intimate knowledge of RDF *and nobody else*. If you're not creating tools, use the tools, Luke. -- Julian Bond Email&MSM: julian.bond at voidstar.com Webmaster: http://www.ecademy.com/ Personal WebLog: http://www.voidstar.com/ M: +44 (0)77 5907 2173 T: +44 (0)192 0412 433 From reto at gmuer.ch Tue Sep 14 12:57:57 2004 From: reto at gmuer.ch (Reto Bachmann-Gmuer) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] Role-Vocab and Re: Postal address vocabulary In-Reply-To: References: Message-ID: <4146EAD5.3000908@gmuer.ch> David Menendez wrote: > Graham Klyne writes: > > >>I acknowledge all the points you make in your response. My question >>(and this is a question, not an assertion rhetorically styled as a >>question) is this: should this kind of detailed address-description >>vocabulary be part of FOAF, or a separate vocabulary (e.g. another >>namespace, which could, of course, be used *with* FOAF)? > > > I'd make it a separate vocabulary (call it an adjunct or a satellite > vocabulary). FOAF and the address vocab are both large enough that > they'd be easier to manage separately, and making the address vocab > separate might make it more attractive to other projects. > > The only negative point which comes to mind is that it's slightly more > work for people generating FOAF profiles by hand. I'd suggest to have the address vocab a sperate vocabulary except maybe the #address property which has domain foaf:Agent. Have you had a look at the role vocab at http://rdfweb.org/topic/RoleVocab? I think this could usefully be part of the foaf-core vocab: seing foaf primarily as a vocabulary for describing (social) agents and their relations. Possibly this would make foaf:workplaceHompage obsolete. I've put an example on the wiki page showing a more comprehensive version of the workplaceHomepage example from the foaf-spec. cheers, reto From zednenem at psualum.com Wed Sep 15 01:13:16 2004 From: zednenem at psualum.com (David Menendez) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] Postal address vocabulary In-Reply-To: Message-ID: Julian Bond writes: > Graham Klyne wrote: > >I acknowledge all the points you make in your response. My question > >(and this is a question, not an assertion rhetorically styled as a > >question) is this: should this kind of detailed address-description > >vocabulary be part of FOAF, or a separate vocabulary (e.g. another > >namespace, which could, of course, be used *with* FOAF)? > > I consider a generalised RDF (and/or XML) Address schema to be so > important that it's development *must* be done in the context of a > major standards organization. Given the history, rfc status and wide > deployment and implementation of vCard I find it pretty hard to argue > for a new one. So maybe what's really needed is a new vCard RFC that > supersedes RFC 2426. If the problem is with the RDF representation of > vCard, then that's being done within the W3C. If it's wrong there are > well established ways to improve it. Without agreeing or disagreeing, I'll note: (1) The only RDF schema for vCard I can find is a W3C Technical Note[1] from 2001 (which, in my opinion, could use some refactoring) (2) Reto's proposal[2] is an RDF translation of S42-3 from the Univeral Postal Union[3]. (3) S42-3 deals specifically with postal addresses (it represents the information on an individual envelopes), while vCard has a wider focus (it represents the information one might have on a business card). > David Menendez wrote: > >The only negative point which comes to mind is that it's slightly > >more work for people generating FOAF profiles by hand. > > The only people generating FOAF by hand should be experimenters and > programmer/implementors with an intimate knowledge of RDF *and nobody > else*. If you're not creating tools, use the tools, Luke. Exactly. My point was that the only argument I could think of for adding a full-blown address vocabulary to FOAF was very weak. [1] [2] [3] -- David Menendez From mcculley at stackframe.com Wed Sep 15 04:15:58 2004 From: mcculley at stackframe.com (Gene McCulley) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] www.foafspace.com Message-ID: RDF folks, I have built a foaf search engine and put it at www.foafspace.com. I have not yet finished the smushing mechanism, so the engine currently returns more entries than desired. It still has some problems with the occasional RDF file. Other than that, it is mostly stable. I would appreciate any feedback regarding desired features and bugs found. Thanks. -------------- next part -------------- A non-text attachment was scrubbed... Name: PGP.sig Type: application/pgp-signature Size: 186 bytes Desc: This is a digitally signed message part Url : http://lists.usefulinc.com/cgi-bin/mailman/private/foaf-dev/attachments/20040915/306dcc2b/PGP.pgp From julian_bond at voidstar.com Wed Sep 15 06:51:22 2004 From: julian_bond at voidstar.com (Julian Bond) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] www.foafspace.com In-Reply-To: References: Message-ID: Gene McCulley wrote: >I have built a foaf search engine and put it at www.foafspace.com. > >I have not yet finished the smushing mechanism, so the engine currently >returns more entries than desired. It still has some problems with the >occasional RDF file. Other than that, it is mostly stable. > >I would appreciate any feedback regarding desired features and bugs >found. Nice. What's the underlying technology? And what Schemas does it understand? -- Julian Bond Email&MSM: julian.bond at voidstar.com Webmaster: http://www.ecademy.com/ Personal WebLog: http://www.voidstar.com/ M: +44 (0)77 5907 2173 T: +44 (0)192 0412 433 From john.breslin at deri.org Wed Sep 15 09:46:17 2004 From: john.breslin at deri.org (John Breslin) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] www.foafspace.com In-Reply-To: Message-ID: <41481D79.1154.4DF07A@localhost> Looks cool, just submitted my FOAF profile from boards.jp and it spidered 364 other members from that. Are you just using the unofficial foaf:dateOfBirth or any other variations? J. -- Dr. John Breslin Digital Enterprise Research Institute http://www.johnbreslin.com/ john.breslin@deri.org From mcculley at stackframe.com Wed Sep 15 12:56:17 2004 From: mcculley at stackframe.com (Gene McCulley) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] www.foafspace.com In-Reply-To: References: Message-ID: It uses a Java web application engine that I built for other purposes. The spidering and database management are done in Java as well. I wrote my own library for doing FOAF/RDF parsing, which might not be the right thing to do. I intend to take a look at Jena to see if I should use that for the RDF bits. It currently understands most of the foaf schema. I intend to add geo and vcard support. On Sep 15, 2004, at 2:51 AM, Julian Bond wrote: > Nice. What's the underlying technology? And what Schemas does it > understand? -------------- next part -------------- A non-text attachment was scrubbed... Name: PGP.sig Type: application/pgp-signature Size: 186 bytes Desc: This is a digitally signed message part Url : http://lists.usefulinc.com/cgi-bin/mailman/private/foaf-dev/attachments/20040915/8b667645/PGP.pgp From mcculley at stackframe.com Wed Sep 15 13:05:11 2004 From: mcculley at stackframe.com (Gene McCulley) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] www.foafspace.com In-Reply-To: <41481D79.1154.4DF07A@localhost> References: <41481D79.1154.4DF07A@localhost> Message-ID: Thanks for submitting the boards.jp link. If anyone has pointers to other sites that have multiple FOAF entries (that aren't shown on http://www.foafspace.com/hosts), I would appreciate it. I am using the unofficial dateOfBirth just because that is what LiveJournal emits. I tried to stay away from the bits of the schema not marked stable, but there is a lot of content out there using unofficial tags. I see that some files use the vCard BDAY or a birthday field in the bio schema. I intend to add support for those. On Sep 15, 2004, at 5:46 AM, John Breslin wrote: > Looks cool, just submitted my FOAF profile from boards.jp and it > spidered 364 > other members from that. > > Are you just using the unofficial foaf:dateOfBirth or any other > variations? > > J. > -- > Dr. John Breslin > Digital Enterprise Research Institute > http://www.johnbreslin.com/ > john.breslin@deri.org > > -------------- next part -------------- A non-text attachment was scrubbed... Name: PGP.sig Type: application/pgp-signature Size: 186 bytes Desc: This is a digitally signed message part Url : http://lists.usefulinc.com/cgi-bin/mailman/private/foaf-dev/attachments/20040915/47a5153b/PGP.pgp From richard at cyganiak.de Wed Sep 15 21:57:50 2004 From: richard at cyganiak.de (Richard Cyganiak) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] Issue: FOAF RDFS should not import RDFS and OWL vocabularies Message-ID: <49882A1F-0762-11D9-9A5B-000A95DABFDC@cyganiak.de> Hi everyone, the FOAF RDFS file[1] imports the RDFS and OWL vocabulary files: The OWL guide[2] recommends against doing that: Note that in order to use the OWL vocabulary you do not need to import the owl.rdf ontology. In fact, such an import is not recommended. I've added this to the IssueTracker[3]. It's not a big issue, but I thought I'd mention it anyway. Lots of ontologies/vocabularies do that. Is there a good reason that I am missing? Richard [1] http://xmlns.com/foaf/0.1/index.rdf [2] http://www.w3.org/TR/owl-guide/ [3] http://rdfweb.org/topic/OwlImportsIssue From danbri at w3.org Wed Sep 15 23:46:11 2004 From: danbri at w3.org (Dan Brickley) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] Issue: FOAF RDFS should not import RDFS and OWL vocabularies In-Reply-To: <49882A1F-0762-11D9-9A5B-000A95DABFDC@cyganiak.de> References: <49882A1F-0762-11D9-9A5B-000A95DABFDC@cyganiak.de> Message-ID: <20040915234610.GA18782@homer.w3.org> * Richard Cyganiak [2004-09-15 23:57+0200] > Hi everyone, > > the FOAF RDFS file[1] imports the RDFS and OWL vocabulary files: > > > > > The OWL guide[2] recommends against doing that: > > Note that in order to use the OWL vocabulary you do not need to > import the owl.rdf ontology. In fact, such an import is not > recommended. > > I've added this to the IssueTracker[3]. It's not a big issue, but I > thought I'd mention it anyway. Lots of ontologies/vocabularies do that. > Is there a good reason that I am missing? I added the owl:imports statements (with some skepticism) pretty early on in OWL's lifetime, I think (FOAF predates OWL). I'd be quite happy to remove these statements, and perhaps just add rdf:seeAlso crossrefs to provide a hint for crawlers (since multilingual labels can be found via crawling from the rdfs namespace). owl:imports is imho a bit wierd, sort of mixes up the layers in a strange way, and doesn't add much anyway... If anyone objects to removing these triples, now would be a great time to say! Dan > > Richard > > [1] http://xmlns.com/foaf/0.1/index.rdf > [2] http://www.w3.org/TR/owl-guide/ > [3] http://rdfweb.org/topic/OwlImportsIssue > > > _______________________________________________ > rdfweb-dev mailing list > rdfweb-dev@vapours.rdfweb.org > wiki: http://rdfweb.org/topic/FoafProject > http://rdfweb.org/mailman/listinfo/rdfweb-dev From ronwalf at volus.net Thu Sep 16 01:30:14 2004 From: ronwalf at volus.net (Ron Alford) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] Issue: FOAF RDFS should not import RDFS and OWL vocabularies In-Reply-To: <20040915234610.GA18782@homer.w3.org> References: <49882A1F-0762-11D9-9A5B-000A95DABFDC@cyganiak.de> <20040915234610.GA18782@homer.w3.org> Message-ID: <4148ECA6.3030002@volus.net> Dan Brickley wrote: > I added the owl:imports statements (with some skepticism) pretty early > on in OWL's lifetime, I think (FOAF predates OWL). I'd be quite happy to > remove these statements, and perhaps just add rdf:seeAlso crossrefs to > provide a hint for crawlers (since multilingual labels can be found via > crawling from the rdfs namespace). Yeah, a seeAlso would be fine there, and probably more along the lines of what you mean. > owl:imports is imho a bit wierd, sort of mixes up the layers in a > strange way, and doesn't add much anyway... owl:imports is only really needed at the ontology level if you are really making use of the structural parts of another ontology. What it's really useful for is in instance data. I use it over at http://www.mindswap.org/2004/owl/mindswappers to point to an organization-specific extension of foaf (which happens to point to a dl-friendly version of the foaf ont). The owl:imports has well defined semantics with respect to reasoning. This makes it easy for a spidering owl agent to perform inferences over a file without having to already know about all the schemas the data could be using. > > If anyone objects to removing these triples, now would be a great time > to say! > Removing them would be great! I'm glad to see some interest from someone other than myself about making foaf more owl-friendly! -Ron From Libby.Miller at bristol.ac.uk Thu Sep 16 09:25:18 2004 From: Libby.Miller at bristol.ac.uk (Libby Miller) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] Issue: FOAF RDFS should not import RDFS and OWL vocabularies In-Reply-To: <49882A1F-0762-11D9-9A5B-000A95DABFDC@cyganiak.de> References: <49882A1F-0762-11D9-9A5B-000A95DABFDC@cyganiak.de> Message-ID: Thanks for this. I think someone else mentionned it (possibly in conversation) so it's good to have it documented. I'm not sure why the issue is important, though... Libby On Wed, 15 Sep 2004, Richard Cyganiak wrote: > Hi everyone, > > the FOAF RDFS file[1] imports the RDFS and OWL vocabulary files: > > > > > The OWL guide[2] recommends against doing that: > > Note that in order to use the OWL vocabulary you do not need to > import the owl.rdf ontology. In fact, such an import is not > recommended. > > I've added this to the IssueTracker[3]. It's not a big issue, but I thought > I'd mention it anyway. Lots of ontologies/vocabularies do that. Is there a > good reason that I am missing? > > Richard > > [1] http://xmlns.com/foaf/0.1/index.rdf > [2] http://www.w3.org/TR/owl-guide/ > [3] http://rdfweb.org/topic/OwlImportsIssue > > > _______________________________________________ > rdfweb-dev mailing list > rdfweb-dev@vapours.rdfweb.org > wiki: http://rdfweb.org/topic/FoafProject > http://rdfweb.org/mailman/listinfo/rdfweb-dev > From jo at abduction.org Thu Sep 16 15:11:12 2004 From: jo at abduction.org (Jo Walsh) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] Re: [Geowanking] next gen geoURL In-Reply-To: References: <6.0.0.22.2.20040916120413.01b3b168@sunrise.sli.unimelb.edu.au> <753344F9-078F-11D9-96C4-000A95DA2B18@eogeo.org> <453B13F6-0791-11D9-8346-000A95C54164@splattercast.com> Message-ID: <20040916151112.GC23473@vishnu.tridity.org> hey allan, geowankers, On Wed, Sep 15, 2004 at 11:52:49PM -0400, Allan Doyle wrote: > For all the people who are chafing because what's out there is not > quite sufficient in some dimension. One of my particular goals is the > ability to put metadata about geographic data into RSS so small > organizations can advertise their data, or can advertise their need for > data. Aggregators can pick those things up and start making matches > among the haves and the have nots. i wrote a quick braindump, couple of months back, for the craigslist techs about geoannotating rss, html etc with rdf. might be of relevance. http://frot.org/geo/craigslist.html having a URI scheme, or set of them, to approximately indicate locales or places, can be as useful as coordinates and i think will come up a lot during the w3 budapest workshop early next month. this approach to rdf-in-html looks nasty because of the necessity of keeping the XML namespace headers. it's not really human-writable; this discussion seems to be calling for human-writability. in an RSS1 context, not such a problem. i think the w3 is discussing this somewhere, but i'm not sure where. i think mike posted here, at the time, my naive attempts to turn the RDFIG geo namespace into an IETF draft, i never pushed this forward anywhere, but i think it encapsulates the 'decimal degrees in WGS84 for simplicity's sake' argument. http://space.frot.org/draft-geo-draft.html - feedback welcomed. -jo From wolf at bluehands.de Thu Sep 16 16:20:05 2004 From: wolf at bluehands.de (Heiner Wolf) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] Re: [Geowanking] next gen geoURL Message-ID: <11D121AB355B69448D3A9F2132D2A3AA42016B@niobe.BlueHands.de> Hi, do you know the IETF effort there. It is on the IETF standards track. Recent action: "Protocol Action: 'A Presence-based GEOPRIV Location Object Format' to Proposed Standard" Maybe it's useful for FOAF-RDF: See: Title : A Presence-based GEOPRIV Location Object Format Author(s) : J. Peterson Filename : draft-ietf-geopriv-pidf-lo-03.txt Pages : 24 Date : 2004-9-10 This draft is a work item of the Geographic Location/Privacy Working Group of the IETF. This document describes an object format for carrying geographical information on the Internet. This location object extends the Presence Information Data Format (PIDF), which was designed for communicating privacy-sensitive presence information and which has similar properties. A URL for this Internet-Draft is: http://www.ietf.org/internet-drafts/draft-ietf-geopriv-pidf-lo-03.txt hw -- Dr. Klaus H. Wolf bluehands GmbH & Co.mmunication KG http://www.bluehands.de/people/hw +49 (0721) 16108 75 -- Jabber enabled Virtual Presence on the Web: http://www.lluna.de/ Open Source Future History: http://www.galactic-developments.com/ From richard at cyganiak.de Thu Sep 16 17:07:33 2004 From: richard at cyganiak.de (Richard Cyganiak) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] Issue (?): URI of owl:Ontology header in the RDFS spec Message-ID: Hi everyone, I noticed another thing about the FOAF RDFS file. I'm not sure if I have a point this time or if it's just nitpicking. The URL of the RDFS file is http://xmlns.com/foaf/0.1/index.rdf . The owl:Ontology header of this file says: References: <49882A1F-0762-11D9-9A5B-000A95DABFDC@cyganiak.de> <49882A1F-0762-11D9-9A5B-000A95DABFDC@cyganiak.de> Message-ID: <5.1.0.14.2.20040916175728.02aa97b8@127.0.0.1> At 10:25 16/09/04 +0100, Libby Miller wrote: >Thanks for this. I think someone else mentionned it (possibly in >conversation) so it's good to have it documented. I'm not sure why the >issue is important, though... Could it be something to do with this... [[ The meaning of an occurrence of a URI is thus determined by its context, which we take to mean the document in which it appears, plus other documents explicitly mentioned in constructs like the OWL importing mechanism. ]] -- http://www-db.research.bell-labs.com/user/pfps/publications/meaning.pdf ? #g ------------ Graham Klyne For email: http://www.ninebynine.org/#Contact From Libby.Miller at bristol.ac.uk Thu Sep 16 18:04:44 2004 From: Libby.Miller at bristol.ac.uk (Libby Miller) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] Issue (?): URI of owl:Ontology header in the RDFS spec In-Reply-To: References: Message-ID: hia http://xmlns.com/foaf/0.1/ should do content negotiation so you get RDF if you ask for it, html if you don't. Unfortunately I think the stuff you get in rdf from that location has come adrift from the index.rdf version (which is the better version). This needs fixing, sorry. Libby On Thu, 16 Sep 2004, Richard Cyganiak wrote: > Hi everyone, > > I noticed another thing about the FOAF RDFS file. I'm not sure if I have a > point this time or if it's just nitpicking. > > The URL of the RDFS file is http://xmlns.com/foaf/0.1/index.rdf . > > The owl:Ontology header of this file says: > > > The two URIs don't match. When the RDFS file is processed, the processor > cannot know that the statements in the owl:Ontology element (dc:title, > dc:date etc.) are about the current file. The processor will believe that > they refer to another resource located somewhere else. > > The rdf:about URI should be changed, or an xml:base attribute added. Right? > > Thanks, > Richard > > > _______________________________________________ > rdfweb-dev mailing list > rdfweb-dev@vapours.rdfweb.org > wiki: http://rdfweb.org/topic/FoafProject > http://rdfweb.org/mailman/listinfo/rdfweb-dev > From richard at cyganiak.de Thu Sep 16 21:50:06 2004 From: richard at cyganiak.de (Richard Cyganiak) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] Issue (?): URI of owl:Ontology header in the RDFS spec In-Reply-To: References: Message-ID: <5F1B7674-082A-11D9-9A5B-000A95DABFDC@cyganiak.de> Hi Libby, > http://xmlns.com/foaf/0.1/ > > should do content negotiation so you get RDF if you ask for it, html > if you don't. Unfortunately I think the stuff you get in rdf from that > location has come adrift from the index.rdf version (which is the > better version). This needs fixing, sorry. I tried this now, and content negotiation works (Cool!) and the returned version is identical to the index.rdf one. So the only problem is that the URI of the owl:Ontology element does not match the document URL when you fetch from the index.rdf URL. This could be fixed by adding an xml:base="http://xmlns.com/foaf/0.1/", I think. I'll add it to the IssueTracker. Richard From bnowack at appmosphere.com Thu Sep 16 23:32:02 2004 From: bnowack at appmosphere.com (Benjamin Nowack) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] Issue (?): URI of owl:Ontology header in the RDFS spec In-Reply-To: <5F1B7674-082A-11D9-9A5B-000A95DABFDC@cyganiak.de> References: <5F1B7674-082A-11D9-9A5B-000A95DABFDC@cyganiak.de> Message-ID: On 16.09.2004 23:50:06, Richard Cyganiak wrote: > >I tried this now, and content negotiation works (Cool!) and the >returned version is identical to the index.rdf one. So the only problem >is that the URI of the owl:Ontology element does not match the document >URL when you fetch from the index.rdf URL. > >This could be fixed by adding an xml:base="http://xmlns.com/foaf/0.1/", >I think. Hm, I'm not sure if that needs "fixing"... I see the "index.rdf" doc as a copy/mirror of the the ontology doc located at "http://xmlns.com/foaf/0.1/" (via conneg). So the rdf/xml isn't about the "index.rdf" but the spec at the "official" URL. (i.e. you could put a copy on an arbitrary server, e.g. "http://example.com/foaf_spec.rdf" but would probably still keep the rdf:about="http://xmlns.com/foaf/0.1/" unless you wanted to talk about the copied spec.) As the foaf rdfs doc doesn't use rdf:ID for its terms, I don't think that an xml:base is needed. (it could perhaps be a (different) issue that the conneg "feature" leads to an overloaded URI/URL. We can't unambiguously use it to e.g. talk about the html representation vs the rdf/xml version at "http://xmlns.com/foaf/0.1/" ...) good night, benjamin -- Benjamin Nowack Kruppstr. 100 45145 Essen, Germany http://www.appmosphere.com/ > >I'll add it to the IssueTracker. > >Richard > > >_______________________________________________ >rdfweb-dev mailing list >rdfweb-dev@vapours.rdfweb.org >wiki: http://rdfweb.org/topic/FoafProject >http://rdfweb.org/mailman/listinfo/rdfweb-dev > > From saul at twenteenthcentury.com Fri Sep 17 03:09:52 2004 From: saul at twenteenthcentury.com (Saul Albert) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] distributed library project booklists in rdf Message-ID: <20040917030952.GG48246@chinabone.lth.bclub.org.uk> Dear RDFwebs, I've been lurking on this list for the *longest* time.. and I'm very aware of my technical inexperience.. much of it goes way over my head, so apologies if the question is misplaced or just silly: I've been working with a group on the distributed library project software - a books.burri.to style like thing but with less well sructured data, a miserable back end and basically a big pile of ugly hacks in php, but (unlike books.burri.to) people do actually use it to catalogue and share their books and metadata about stuff that otherwise might not be catalogued or accessible: - http://dlp.theps.net - http://dlp.theps.net/maps/ (consume.net style map of library nodes around limehouse town hall - http://twenteenthcentury.com/lth) I decided a while ago that it was beyond my short-term abilities to restructure the data model of the dlp and re-write it all from scratch, especially as people are actually using it. The value of the project then is simply data aggregation - at the moment it does not much with what is a very interesting data set (potentially this could be a way of generating (wikipedia style) the largest private library collection around given that there are now over 26 library nodes in 14 countries). Also, the prospect of semantic links between different-language materials through their cross-associations with other books / book lists etc. is an enticing one. So my question is this: If you were going to produce some interesting apps to scrape lots and lots of very interesting data sets form a bunch of distributed library nodes (basically, bibliographic information plus reviews, del.icio.us like tags (coming soon I hope) and geolocation / ownership / borrowing associations etc..) and do all kinds of magical associcative / locative / meaning-mangling stuff with it, what format would be most useful to you? I'll try to do standard library data exchange export formats for individual's librarys and the whole database of each node, but I'm interested to know if anyone has worked on any rdf that might be suitable for this.. I've seen this http://www.cellml.org/public/metadata/citations.html and jo did some graph models ages ago that would be useful if I could figure out what the foaf file they would come from would look like: http://dlpdev.theps.net/DlpGraphModels Basically I'm hoping someone will have done something similar - bibliographic records cross referenced with geolocation and other associations of ownership, tags etc... that I can just rip off and churn out from our databases.. anyone? thanks Saul. -- other people's sigs experience my ethereal equine art --- http://www.amandakoh.com and help to free art for all --- http://www.prodigalart.org We are all one. -- From stpeter at jabber.org Fri Sep 17 03:52:37 2004 From: stpeter at jabber.org (Peter Saint-Andre) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] full of questions Message-ID: After a hiatus of about a year, I've re-started serious research into integrating FOAF and other RDF vocabularies into Jabber/XMPP. I'm sure I'll have quite a few questions as I move along and work to write a Jabber Enhancement Proposal that defines the integration. Here are some starter questions: 1. Is it possible to define oneself as a member of a group? I like the fact that a Group is its own kind of Agent, but I think it would be helpful to be able to say "I'm a member of the group 'JSF Members' defined by http://www.jabber.org/members/foaf.rdf" (no FOAF file at that URL yet, but I'll make one soon). I realize that in true RDF fashion we'd expect the appropriate software agents to crawl around and find that group definition, then connect me with that group if its FOAF file says I am a member. Is there no shortcut like foaf:memberOf or something like that? 2. Connected to this: is there a property that enables me to say "I subscribe to this mailing list" or something of that kind? Is that simply an interest? We might not want to publish this Group, i.e., the list of folks who subscribe to a list (in fact as a list admin I always set access to the subscriber list to "list admin only" in Mailman), but I might want to identify myself as a subscriber to a list. I suppose making this an interest might work. 3. Are there any properties for things like marital status and other dating-related items (e.g., all the usual TLAs one sees in personals ads)? As you can imagine, such attributes are of more than passing interest in the IM world. 4. Did you folks ever settle any of the issues surrounding the representation of physical addresses? We're working to fully replace everything that currently exists in vCard, so address representation is necessary for us. 5. On the Jabber network, we might want to create FOAF representations for things like chatrooms, bots, and IM servers (currently many Jabber servers have their own vCards). We could see a chatroom as a kind of Group, but a bot is not a Person (though it is an Agent) and I would characterize a server as an AccountService of some kind (I don't see how it fits into any of the existing Agent categories). 6. An Internet-Draft defining the xmpp: URI scheme is pretty far along within the IETF, so it seems we might want to use that in representing the foaf:jabberID property. (Then again, we might not, given the wild escaping rules and such for URIs; a jabberID is at root a UTF-8 string that allows all sorts of nice internationalized support, whereas a URI needs to escape non-US-ASCII characters; so perhaps leaving things as they are would be better -- but at the least we need to specify whether a foaf:jabberID is an XMPP URI or a native XMPP address.) 7. Regarding the OnlineChatAccount property, I am wondering what would be appropriate for the accountServiceHomepage. As you may recall, the Jabber network has a distributed architecture, and anyone may run their own Jabber server (just as they may do with email). So it seems that each Jabber server would have its own homepage and that we can't define one generic URI here as might be able to do with Freenode (as in the example shown in the FOAF spec), ICQ, AIM, and so on. I'd like to help define the mapping between the abbreviated foaf:jabberID property and the full OnlineChatAccount representation, so we'd need to figure this out in order to make that happen. I'm happy to work on defining any of the attributes we're looking for that don't exist yet. More to come, I'm sure... Peter From Libby.Miller at bristol.ac.uk Fri Sep 17 04:56:48 2004 From: Libby.Miller at bristol.ac.uk (Libby Miller) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] Issue (?): URI of owl:Ontology header in the RDFS spec In-Reply-To: <5F1B7674-082A-11D9-9A5B-000A95DABFDC@cyganiak.de> References: <5F1B7674-082A-11D9-9A5B-000A95DABFDC@cyganiak.de> Message-ID: Thanks Richard! Libby On Thu, 16 Sep 2004, Richard Cyganiak wrote: > Hi Libby, > >> http://xmlns.com/foaf/0.1/ >> >> should do content negotiation so you get RDF if you ask for it, html if you >> don't. Unfortunately I think the stuff you get in rdf from that location >> has come adrift from the index.rdf version (which is the better version). >> This needs fixing, sorry. > > I tried this now, and content negotiation works (Cool!) and the returned > version is identical to the index.rdf one. So the only problem is that the > URI of the owl:Ontology element does not match the document URL when you > fetch from the index.rdf URL. > > This could be fixed by adding an xml:base="http://xmlns.com/foaf/0.1/", I > think. > > I'll add it to the IssueTracker. > > Richard > > From danbri at w3.org Fri Sep 17 08:35:46 2004 From: danbri at w3.org (Dan Brickley) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] Issue (?): URI of owl:Ontology header in the RDFS spec In-Reply-To: References: <5F1B7674-082A-11D9-9A5B-000A95DABFDC@cyganiak.de> Message-ID: <20040917083546.GA7435@homer.w3.org> * Benjamin Nowack [2004-09-17 01:32+0200] > On 16.09.2004 23:50:06, Richard Cyganiak wrote: > > > >I tried this now, and content negotiation works (Cool!) and the > >returned version is identical to the index.rdf one. So the only problem > >is that the URI of the owl:Ontology element does not match the document > >URL when you fetch from the index.rdf URL. > > > >This could be fixed by adding an xml:base="http://xmlns.com/foaf/0.1/", > >I think. > Hm, I'm not sure if that needs "fixing"... > I see the "index.rdf" doc as a copy/mirror of the the ontology > doc located at "http://xmlns.com/foaf/0.1/" (via conneg). So the rdf/xml > isn't about the "index.rdf" but the spec at the "official" URL. > (i.e. you could put a copy on an arbitrary server, e.g. > "http://example.com/foaf_spec.rdf" but would probably still keep the > rdf:about="http://xmlns.com/foaf/0.1/" unless you wanted to talk about > the copied spec.) > As the foaf rdfs doc doesn't use rdf:ID for its terms, I don't think > that an xml:base is needed. That's how I see it too. The index.rdf is a convenience resource. There's also a copy embedded raw in the xhtml version, as well as content negotiable. A 3rd way to see it was as entity-escaped colourised markup (generated with XSLT), but I removed that as the generator script was broken and leaving the human-visible markup appendix unreadable. It also made the spec really long. The markup was designed so it'd generated the same triples. Also btw should work so that archival copies of the namespace could be made, eg. 20040901-index.rdf or whatever, and generate triples that provide a 'historical' view that talks about the same things, rather than talking about properties/classes whose URIs contain '20040901' etc... For all that, it should be harmless to include the xml:base declaration, might make the intent clearer. Dan > > (it could perhaps be a (different) issue that the conneg "feature" > leads to an overloaded URI/URL. We can't unambiguously use it to e.g. > talk about the html representation vs the rdf/xml version at > "http://xmlns.com/foaf/0.1/" ...) > > good night, > benjamin > > -- > Benjamin Nowack > > Kruppstr. 100 > 45145 Essen, Germany > http://www.appmosphere.com/ > > > > >I'll add it to the IssueTracker. > > > >Richard > > > > > >_______________________________________________ > >rdfweb-dev mailing list > >rdfweb-dev@vapours.rdfweb.org > >wiki: http://rdfweb.org/topic/FoafProject > >http://rdfweb.org/mailman/listinfo/rdfweb-dev > > > > > > > _______________________________________________ > rdfweb-dev mailing list > rdfweb-dev@vapours.rdfweb.org > wiki: http://rdfweb.org/topic/FoafProject > http://rdfweb.org/mailman/listinfo/rdfweb-dev From rich.boakes at port.ac.uk Fri Sep 17 09:11:06 2004 From: rich.boakes at port.ac.uk (rich boakes) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] Issue (?): URI of owl:Ontology header in the RDFS spec In-Reply-To: <20040917083546.GA7435@homer.w3.org> References: <5F1B7674-082A-11D9-9A5B-000A95DABFDC@cyganiak.de> <20040917083546.GA7435@homer.w3.org> Message-ID: <414AAA2A.5050009@port.ac.uk> danbri wrote: > A 3rd way to see it was as entity-escaped colourised > markup (generated with XSLT), but I removed that as > the generator script was broken ... He Dan, Does the XSLT part still exist and was it at a level where it might be generally usable? Rich From danbri at w3.org Fri Sep 17 09:19:30 2004 From: danbri at w3.org (Dan Brickley) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] Issue (?): URI of owl:Ontology header in the RDFS spec In-Reply-To: <414AAA2A.5050009@port.ac.uk> References: <5F1B7674-082A-11D9-9A5B-000A95DABFDC@cyganiak.de> <20040917083546.GA7435@homer.w3.org> <414AAA2A.5050009@port.ac.uk> Message-ID: <20040917091930.GI7435@homer.w3.org> * rich boakes [2004-09-17 10:11+0100] > danbri wrote: > > A 3rd way to see it was as entity-escaped colourised > > markup (generated with XSLT), but I removed that as > > the generator script was broken ... > > He Dan, > > Does the XSLT part still exist and was it at a level where > it might be generally usable? The XSLT is from Max Froumentin, see http://www.w3.org/People/maxf/ under "XML file pretty printer" http://www.w3.org/2003/02/colour-xml-serializer.xsl I think it needs a few args for setting the colours, and can't in XSLT 1 be entirely general purpose. But the code's up there... cheers, Dan From richard at cyganiak.de Fri Sep 17 10:58:03 2004 From: richard at cyganiak.de (Richard Cyganiak) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] Issue (?): URI of owl:Ontology header in the RDFS spec In-Reply-To: <20040917083546.GA7435@homer.w3.org> References: <5F1B7674-082A-11D9-9A5B-000A95DABFDC@cyganiak.de> <20040917083546.GA7435@homer.w3.org> Message-ID: <72D47006-0898-11D9-9A5B-000A95DABFDC@cyganiak.de> Hi Dan, hi Benjamin, thanks for your answers. Am 17.09.2004 um 10:35 schrieb Dan Brickley: > * Benjamin Nowack [2004-09-17 01:32+0200] >> On 16.09.2004 23:50:06, Richard Cyganiak wrote: >>> the URI of the owl:Ontology element does not match the document >>> URL when you fetch from the index.rdf URL. >>> >>> This could be fixed by adding an >>> xml:base="http://xmlns.com/foaf/0.1/", >>> I think. >> Hm, I'm not sure if that needs "fixing"... > That's how I see it too. The index.rdf is a convenience resource. Here is why I think the document currently served from .../index.rdf is broken: Imagine I want to open the FOAF spec in my RDFS/OWL ontology editor. So I google for "foaf rdfs" or search the HTML spec for a link to the RDFS and end up with the .../index.rdf link. I open it in my editor. All the classes and properties show up just fine. But there's no metadata about the ontology itself, like title, date and so on, like you have in pretty much every RDFS file you can find on the net. So they didn't include that in the FOAF RDFS? How lazy! On closer inspection I see that the info is there in the file, but my ontology editor didn't show it because it was unable to associate the owl:Ontology header with the current document. The editor "thought" that, for some reason, the file contained metadata about some other ontology somewhere on the Web with no apparent relation to the current file. Needs fixing or not, from that point of view? I absolutely agree with all your good points about index.rdf just being a mirror, archival copies, and having the same triples in all versions. None of these would be harmed by an xml:base declaration. Benjamin: >> As the foaf rdfs doc doesn't use rdf:ID for its terms, I don't think >> that an xml:base is needed. As I see it, xml:base is used for three things: 1. building URIs from rdf:ID, 2. building absolute URIs from relative URIs, as in rdf:about="#foo". 3. determining what the URI of the document itself is, as in rdf:about="". 1 and 2 are not relevant for the FOAF RDFS. 3 is. Dan: > For all that, it should be harmless to include the xml:base > declaration, > might make the intent clearer. That would be great! Thanks, Richard From danbri at w3.org Fri Sep 17 11:09:06 2004 From: danbri at w3.org (Dan Brickley) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] Issue (?): URI of owl:Ontology header in the RDFS spec In-Reply-To: <72D47006-0898-11D9-9A5B-000A95DABFDC@cyganiak.de> References: <5F1B7674-082A-11D9-9A5B-000A95DABFDC@cyganiak.de> <20040917083546.GA7435@homer.w3.org> <72D47006-0898-11D9-9A5B-000A95DABFDC@cyganiak.de> Message-ID: <20040917110906.GM7435@homer.w3.org> * Richard Cyganiak [2004-09-17 12:58+0200] > Hi Dan, hi Benjamin, > > thanks for your answers. > > Am 17.09.2004 um 10:35 schrieb Dan Brickley: > >* Benjamin Nowack [2004-09-17 01:32+0200] > >>On 16.09.2004 23:50:06, Richard Cyganiak wrote: > >>>the URI of the owl:Ontology element does not match the document > >>>URL when you fetch from the index.rdf URL. > >>> > >>>This could be fixed by adding an > >>>xml:base="http://xmlns.com/foaf/0.1/", > >>>I think. > >>Hm, I'm not sure if that needs "fixing"... > > >That's how I see it too. The index.rdf is a convenience resource. > > > Here is why I think the document currently served from .../index.rdf is > broken: > > Imagine I want to open the FOAF spec in my RDFS/OWL ontology editor. So > I google for "foaf rdfs" or search the HTML spec for a link to the RDFS > and end up with the .../index.rdf link. I open it in my editor. All the > classes and properties show up just fine. But there's no metadata about > the ontology itself, like title, date and so on, like you have in > pretty much every RDFS file you can find on the net. So they didn't > include that in the FOAF RDFS? How lazy! > > On closer inspection I see that the info is there in the file, but my > ontology editor didn't show it because it was unable to associate the > owl:Ontology header with the current document. The editor "thought" > that, for some reason, the file contained metadata about some other > ontology somewhere on the Web with no apparent relation to the current > file. > > Needs fixing or not, from that point of view? I'm not familiar enough with the behaviour of the various ontology editing tools out there, but it does sound like an xml:base would help them out. And it should be harmless for other purposes, so I'll add it. In the Semantic Web Best Practices WG at W3C we are just starting to examine such issues within a new 'Vocabulary Management' Task Force. I'm hoping FOAF can be made into a case study area there, so I expect to come back to this issue later, and see how other namespaces are managing this problem. In meantime, I'll add xml:base as part of next update to the site. cheerrs, Dan > I absolutely agree with all your good points about index.rdf just being > a mirror, archival copies, and having the same triples in all versions. > None of these would be harmed by an xml:base declaration. > > > Benjamin: > >>As the foaf rdfs doc doesn't use rdf:ID for its terms, I don't think > >>that an xml:base is needed. > > As I see it, xml:base is used for three things: > > 1. building URIs from rdf:ID, > 2. building absolute URIs from relative URIs, as in rdf:about="#foo". > 3. determining what the URI of the document itself is, as in > rdf:about="". > > 1 and 2 are not relevant for the FOAF RDFS. 3 is. > > > Dan: > >For all that, it should be harmless to include the xml:base > >declaration, > >might make the intent clearer. > > That would be great! > > Thanks, > Richard From bnowack at appmosphere.com Fri Sep 17 12:12:52 2004 From: bnowack at appmosphere.com (Benjamin Nowack) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] Issue (?): URI of owl:Ontology header in the RDFS spec In-Reply-To: <72D47006-0898-11D9-9A5B-000A95DABFDC@cyganiak.de> References: <5F1B7674-082A-11D9-9A5B-000A95DABFDC@cyganiak.de> <20040917083546.GA7435@homer.w3.org> <72D47006-0898-11D9-9A5B-000A95DABFDC@cyganiak.de> Message-ID: On 17.09.2004 12:58:03, Richard Cyganiak wrote: >Here is why I think the document currently served from .../index.rdf is >broken: > >Imagine I want to open the FOAF spec in my RDFS/OWL ontology editor. So >I google for "foaf rdfs" or search the HTML spec for a link to the RDFS >and end up with the .../index.rdf link. I open it in my editor. All the >classes and properties show up just fine. But there's no metadata about >the ontology itself, like title, date and so on, like you have in >pretty much every RDFS file you can find on the net. So they didn't >include that in the FOAF RDFS? How lazy! > >On closer inspection I see that the info is there in the file, but my >ontology editor didn't show it because it was unable to associate the >owl:Ontology header with the current document. The editor "thought" >that, for some reason, the file contained metadata about some other >ontology somewhere on the Web with no apparent relation to the current >file. > >Needs fixing or not, from that point of view? if your ontology editor really behaves that way, then I guess it'd need fixing, too ;) afaik, none of the resources described in the foaf spec are related to the "index.rdf" url. all the terms are identified by *absolute* uris, e.g. rdf:about="http://xmlns.com/foaf/0.1/Person". And so is the ontology. Why should your editor read the terms but not the ontology resource? And adding an xml:base="http://xmlns.com/foaf/0.1/" wouldn't actually change the ontology's uri as it doesn't have a relative uri assigned. either version leads to exactly the same triples. But I agree with Dan, adding the base doesn't hurt. >I absolutely agree with all your good points about index.rdf just being >a mirror, archival copies, and having the same triples in all versions. >None of these would be harmed by an xml:base declaration. see above, true. see below, maybe false. >Benjamin: >>> As the foaf rdfs doc doesn't use rdf:ID for its terms, I don't think >>> that an xml:base is needed. > >As I see it, xml:base is used for three things: > >1. building URIs from rdf:ID, >2. building absolute URIs from relative URIs, as in rdf:about="#foo". >3. determining what the URI of the document itself is, as in >rdf:about="". > >1 and 2 are not relevant for the FOAF RDFS. 3 is. but there is no rdf:about="" in either of "/" or "index.rdf" docs, or am I looking at the wrong docs? In the latter case I absolutely agree with you, but I get both times here. And to be even more nitpicking (sorry ;), *if* we add an xml:base to the spec doc at index.rdf, we couldn't easily add an additional which would make it possible to describe circulated copies. Imagine we had a tool that created a copy of the spec every month to keep those "snapshots" dan mentioned in the wiki. It would be easy to add a dc:date to the additional ontology header and then save this doc whereever you want. the dc:date would always point to the specific copy. (well, this doesn't help with the terms' absolute uris, if those copies were published, but still..) ;) benjamin -- Benjamin Nowack Kruppstr. 100 45145 Essen, Germany http://www.appmosphere.com/ From stpeter at jabber.org Fri Sep 17 15:33:37 2004 From: stpeter at jabber.org (Peter Saint-Andre) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] QuoteIssue Message-ID: I have created a QuoteIssue in the IssueTracker: http://rdfweb.org/topic/QuoteIssue Abstract: The recent Pew Internet survey on instant messaging noted that the most popular form of content added to IM profile pages (think the ICQ website) is "funny quotes or other sayings" that the person identifies with in some way. The numbers: 42% of profile creators included quotes, whereas only 18% included links to websites of interest (think foaf:interest). So it seems that it would be good for FOAF to include a foaf:quote property of some kind. Let me know if more detailed information is desired, and do tell me how to proceed with this. Shall I write up some proposed text and examples? I am happy to be the issue owner for this. Peter From stpeter at jabber.org Fri Sep 17 15:34:42 2004 From: stpeter at jabber.org (Peter Saint-Andre) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] AccountServiceIssue Message-ID: I have created an AccountServiceIssue in the IssueTracker: http://rdfweb.org/topic/AccountServiceIssue Abstract: On the Jabber network, we might want to create FOAF representations for things like chatrooms, bots, and IM servers. In particular, currently many Jabber servers have their own vCards and we would want to replace those with a FOAF representation. I think that an IM server is a kind of Agent that is not yet discussed elsewhere, which I would characterize as an AccountService. This terms needs further definition but I think it captures the core idea (related, naturally, to the existing accountServiceHomepage property). Let me know if more detailed information is desired, and do tell me how to proceed with this. Shall I write up some proposed text and examples? I am happy to be the issue owner for this. Peter From richard at cyganiak.de Fri Sep 17 18:26:29 2004 From: richard at cyganiak.de (Richard Cyganiak) Date: Tue Dec 12 14:55:26 2006 Subject: OT: base URIs and semantics (was: Re: [rdfweb-dev] Issue (?): URI of owl:Ontology header in the RDFS spec) In-Reply-To: References: <5F1B7674-082A-11D9-9A5B-000A95DABFDC@cyganiak.de> <20040917083546.GA7435@homer.w3.org> <72D47006-0898-11D9-9A5B-000A95DABFDC@cyganiak.de> Message-ID: <17A7E7E6-08D7-11D9-9A5B-000A95DABFDC@cyganiak.de> Hi Benjamin, to keep this short: I believe that a document's base URI is part of its semantics. Change the base URI, change the semantics. This is not stated in any spec, but it's a way of thinking that works quite well for me. Triples are not everything. There are containers for triples too. rdf:about="" and rdf:about="$baseURI" are ways to say things about containers for triples. Remember this when moving these containers around. xml:base is a way to make sure you can move these containers around without breaking any statements about them. But, granted, this is one of those "Here be dragons" areas of RDF semantics, so you are free to disagree, and I can't prove wrong anything you said below. Richard > On 17.09.2004 12:58:03, Richard Cyganiak wrote: >> Here is why I think the document currently served from .../index.rdf >> is >> broken: >> >> Imagine I want to open the FOAF spec in my RDFS/OWL ontology editor. >> So >> I google for "foaf rdfs" or search the HTML spec for a link to the >> RDFS >> and end up with the .../index.rdf link. I open it in my editor. All >> the >> classes and properties show up just fine. But there's no metadata >> about >> the ontology itself, like title, date and so on, like you have in >> pretty much every RDFS file you can find on the net. So they didn't >> include that in the FOAF RDFS? How lazy! >> >> On closer inspection I see that the info is there in the file, but my >> ontology editor didn't show it because it was unable to associate the >> owl:Ontology header with the current document. The editor "thought" >> that, for some reason, the file contained metadata about some other >> ontology somewhere on the Web with no apparent relation to the current >> file. >> >> Needs fixing or not, from that point of view? > if your ontology editor really behaves that way, then I guess it'd need > fixing, too ;) > afaik, none of the resources described in the foaf spec are related > to the "index.rdf" url. all the terms are identified by *absolute* > uris, > e.g. rdf:about="http://xmlns.com/foaf/0.1/Person". And so is the > ontology. > Why should your editor read the terms but not the ontology resource? > And > adding an xml:base="http://xmlns.com/foaf/0.1/" wouldn't actually > change > the ontology's uri as it doesn't have a relative uri assigned. either > version leads to exactly the same triples. But I agree with Dan, adding > the base doesn't hurt. > >> I absolutely agree with all your good points about index.rdf just >> being >> a mirror, archival copies, and having the same triples in all >> versions. >> None of these would be harmed by an xml:base declaration. > see above, true. see below, maybe false. > > >> Benjamin: >>>> As the foaf rdfs doc doesn't use rdf:ID for its terms, I don't think >>>> that an xml:base is needed. >> >> As I see it, xml:base is used for three things: >> >> 1. building URIs from rdf:ID, >> 2. building absolute URIs from relative URIs, as in rdf:about="#foo". >> 3. determining what the URI of the document itself is, as in >> rdf:about="". >> >> 1 and 2 are not relevant for the FOAF RDFS. 3 is. > but there is no rdf:about="" in either of "/" or "index.rdf" docs, or > am I looking at the wrong docs? In the latter case I absolutely agree > with you, but I get > > both times here. > > And to be even more nitpicking (sorry ;), *if* we add an xml:base to > the spec doc at index.rdf, we couldn't easily add an additional > which would make it possible to describe > circulated copies. Imagine we had a tool that created a copy of the > spec every month to keep those "snapshots" dan mentioned in the wiki. > It would be easy to add a dc:date to the additional ontology header > and then save this doc whereever you want. the dc:date would always > point to the specific copy. (well, this doesn't help with the terms' > absolute uris, if those copies were published, but still..) > > > ;) > benjamin > > -- > Benjamin Nowack > > Kruppstr. 100 > 45145 Essen, Germany > http://www.appmosphere.com/ > > From danny.ayers at gmail.com Fri Sep 17 20:11:33 2004 From: danny.ayers at gmail.com (Danny Ayers) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] Re: RDF model sync, HTTP deltas, syndicated feeds (and a bounty) In-Reply-To: References: <1f2ed5cd0409130426540c86f9@mail.gmail.com> Message-ID: <1f2ed5cd0409171311669fec64@mail.gmail.com> (thanks stpeter) Update - an entry-oriented diff plugin for Apache has been put together - mod_speedyfeed, see: http://asdf.blogs.com/asdf/2004/09/mod_speedyfeed__1.html some discussion: http://www.intertwingly.net/blog/2004/09/17/mod-speedyfeed I've asked whether it might be suitable for fairly arbitrary RDF/XML - aside from simplifying sync/sharing big graphs, this would be a boon to the presence foafsters, so suggestions appreciated: http://dannyayers.com/archives/2004/09/17/mod_speedyfeed-for-rdfxml/ Cheers, Danny. On Mon, 13 Sep 2004 15:50:11 -0600, Peter Saint-Andre wrote: > In article <1f2ed5cd0409130426540c86f9@mail.gmail.com>, > Danny Ayers wrote: > > > http://bobwyman.pubsub.com/main/2004/09/using_rfc3229_w.html > > Bob also mentions that a true pubsub approach would be best in the long > term; for instance, see: > > http://www.ietf.org/internet-drafts/draft-saintandre-atompub-notify-01.tx > t > > /psa > > _______________________________________________ > rdfweb-dev mailing list > rdfweb-dev@vapours.rdfweb.org > wiki: http://rdfweb.org/topic/FoafProject > http://rdfweb.org/mailman/listinfo/rdfweb-dev > -- http://dannyayers.com From zednenem at psualum.com Sat Sep 18 03:55:14 2004 From: zednenem at psualum.com (David Menendez) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] Issue (?): URI of owl:Ontology header in the RDFS spec In-Reply-To: <72D47006-0898-11D9-9A5B-000A95DABFDC@cyganiak.de> Message-ID: Richard Cyganiak writes: > Here is why I think the document currently served from .../index.rdf > is broken: > > Imagine I want to open the FOAF spec in my RDFS/OWL ontology editor. > So I google for "foaf rdfs" or search the HTML spec for a link to the > RDFS and end up with the .../index.rdf link. I open it in my editor. > All the classes and properties show up just fine. But there's no > metadata about the ontology itself, like title, date and so on, like > you have in pretty much every RDFS file you can find on the net. So > they didn't include that in the FOAF RDFS? How lazy! > > On closer inspection I see that the info is there in the file, but my > ontology editor didn't show it because it was unable to associate the > owl:Ontology header with the current document. The editor "thought" > that, for some reason, the file contained metadata about some other > ontology somewhere on the Web with no apparent relation to the > current file. > > Needs fixing or not, from that point of view? To me, this sounds more like a flaw in the ontology editor. Specifically, it isn't honoring rdfs:isDefinedBy. The FOAF schema has a structure like this (in Turtle): @prefix foaf: . a owl:Ontology. foaf:Agent a owl:Class ; rdfs:isDefinedBy . foaf:Person a owl:Class ; rdfs:isDefinedBy . and so forth. As I see it, the proper behavior for an ontology editor seeing this would be: 1. Find all resources which are instances of owl:Ontology 2. For each ontology, find all resources related to that ontology by rdfs:isDefinedBy. For example, my schema for TDL 3 [1] describes how some of the new terms are related to obsolete terms from earlier versions. An ontology editor would use the rdfs:isDefinedBy relations to learn that tdl:mentions is part of TDL 3, while tdl:commentsOn is not. [1] -- David Menendez | "In this house, we obey the laws | of thermodynamics!" From richard at cyganiak.de Sat Sep 18 11:34:56 2004 From: richard at cyganiak.de (Richard Cyganiak) Date: Tue Dec 12 14:55:26 2006 Subject: OT: rdfs:isDefinedBy and ontology editors (was: Re: [rdfweb-dev] Issue (?): URI of owl:Ontology header in the RDFS spec) In-Reply-To: References: Message-ID: Hi David, Am 18.09.2004 um 05:55 schrieb David Menendez: > To me, this sounds more like a flaw in the ontology editor. > Specifically, it isn't honoring rdfs:isDefinedBy. The semantics of rdfs:isDefinedBy are fuzzy. People might use it for what you suggest, but a couple of different uses would also be covered by the loose definition in the spec. Basing the behaviour of a generic tool on that doesn't seem like a good idea to me. > As I see it, the proper behavior for an ontology editor seeing this > would be: > > 1. Find all resources which are instances of owl:Ontology > 2. For each ontology, find all resources related to that > ontology by rdfs:isDefinedBy. That doesn't work in the common case since both the owl:Ontology header and rdfs:isDefinedBy are optional. > [1] Nice style for the semicolons! Surely makes editing easier. Richard From danny.ayers at gmail.com Sat Sep 18 14:27:03 2004 From: danny.ayers at gmail.com (Danny Ayers) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] TDL and IBIS Message-ID: <1f2ed5cd0409180727197318a2@mail.gmail.com> Hi David, At some point in the not-too-distant future I plan to update the RDF vocabulary [1] I did for IBIS to be more OWLish (IBIS = Issue-Based Information Systems, a collaborative problem analysis/solving technique with a threaded-message model). There's considerable overlap with TDL, so I think it would be worth including equivalentProperty/subPropertyOf etc mappings as appropriate. If and when you have a minute I'd be grateful if you could look over the current schema and see what you reckon. Cheers, Danny. [1] http://purl.org/ibis -- http://dannyayers.com From bnowack at appmosphere.com Tue Sep 21 18:36:52 2004 From: bnowack at appmosphere.com (Benjamin Nowack) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] agenda item suggestion for wednesday's chat Message-ID: hi FOAFers, is the chat going to happen tomorrow (5:30 BST, IIRC)? I'm currently trying to implement some tools related to the FoafCommunityProcess[1]. If the chat is taking place, I'd like to discuss some items of the "Things To Aspire To" section: "identification of term maturity" the spec currently uses "unstable", "testing", "stable" which can be used to describe the current stage of a term's life cycle. If a term is planned to be removed from the term set we could use OWL's deprecation constructs to flag a property or class as deprecated. I wonder if there could be (corner ?) cases where a term's status went from "unstable" to "testing" but after that didn't really get deployed. If the term is not planned to be deprecated, is there a need/way to indicate that this term "was tested, will probably not move to stable but is going to be kept in the spec". This is related to last week's "which terms for names" discussion. Some terms might be in the spec but their use is not recommended any more. Or put in other words: "Tool developers, use these terms. they'll move to 'stable' soon". perhaps this discussion can be replaced by thinking about adding "last modified" or "status changed" annotations to each term. "term documentation" I volunteered at foafcamp to build some kind of a "term browser" for easier access to the description of selected foaf terms. Each term view could show a term's standard information (label, comment, term_status, relations, description, rdf/xml), but also more detailed stuff such as examples, links to tools, use cases, best practices, similar/related terms, version info, issues, todos, etc. Any thoughts on useful documentation types? Just some ideas.. cheers, benjamin [1] http://rdfweb.org/topic/FOAFCommunityProcess -- Benjamin Nowack Kruppstr. 100 45145 Essen, Germany http://www.appmosphere.com/ From tony-rdf at kasei.com Tue Sep 21 22:42:30 2004 From: tony-rdf at kasei.com (Tony Bowden) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] foaf:dateOfBirth In-Reply-To: References: <479B5AE2-E3C7-11D8-9F63-000D932A33F0@atommic.com> Message-ID: <20040921224230.GP8052@soto.kasei.com> On Sun, Aug 01, 2004 at 03:46:31PM +0100, Libby Miller wrote: > You could also use the bio vocab: > [[ > > > 1970-06-15 > Brentwood, Essex, United Kingdom > > > ]] > http://vocab.org/bio/0.1/#Birth Are there other vocabs that include a wider range of events? Rather than just saying that I foaf:knows someone, I'd often prefer to say that I met them at a certain event. What's the preferred way to do something like this? Tony From kjetil at kjernsmo.net Tue Sep 21 23:06:38 2004 From: kjetil at kjernsmo.net (Kjetil Kjernsmo) Date: Tue Dec 12 14:55:26 2006 Subject: Inference about the nature of relationships (was Re: [rdfweb-dev] foaf:dateOfBirth) In-Reply-To: <20040921224230.GP8052@soto.kasei.com> References: <479B5AE2-E3C7-11D8-9F63-000D932A33F0@atommic.com> <20040921224230.GP8052@soto.kasei.com> Message-ID: <200409220106.38501.kjetil@kjernsmo.net> On onsdag 22. september 2004, 00:42, Tony Bowden wrote: > Are there other vocabs that include a wider range of events? Rather > than just saying that I foaf:knows someone, I'd often prefer to say > that I met them at a certain event. What's the preferred way to do > something like this? Eh, well, I had a longish discussion with someone about something like this. Since you brought it up, and I feel I lack understanding, but nevertheless feel that I can elaborate on the problems and on the lack of understanding, I'll follow up anyway. Hope it doesn't sound too stupid to the gurus... :-) Here's my idea: I would like to not only say that I know someone, but why I do know them. Say, for example, about half of the people in my addressbook, I know because they, like me, participate in the sport of orienteering. Furthermore, in my addressbook, I have categorised them as orienteers, and I'd like to express that somehow in my FOAF. Now, one may advocate that just mark yourself up as ed in orienteering. One might then guess that if two people know each other and both share a common interest, they know each other because of that common interest. Well, I have an example where that inference would be wrong: My favorite is actually the even more exotic sport of ski-orienteering, where there are only about 1000 participants in the world. What's the chance of someone sharing interests not knowing eachother because of that...? :-) Well, the other day, it turned out that the guy making a Debian package of GRASS 5.7 is a ski-orienteer. And since I spent less than five hours working with his packages, the chance that I will markup that anytime soon is little, but my interest in ski-orienteering is something that I'll markup as soon as I come up with a good URI for it... I figured maybe the relationship vocab would do this: http://purl.org/vocab/relationship/ particularly the participant(In) properties...? Now, our discussion ended when it was noted that FOAF describes relationships in the form of arcs, and so you have to describe the arc somehow...? Cheers, Kjetil -- Kjetil Kjernsmo Astrophysicist/IT Consultant/Skeptic/Ski-orienteer/Orienteer/Mountaineer kjetil@kjernsmo.net webmaster@skepsis.no editor@learn-orienteering.org Homepage: http://www.kjetil.kjernsmo.net/ OpenPGP KeyID: 6A6A0BBC From danbri at w3.org Wed Sep 22 09:36:47 2004 From: danbri at w3.org (Dan Brickley) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] [collab@sims] CFP: Beyond Personalization 2005 Message-ID: <20040922093647.GD20521@homer.w3.org> It'd be interesting to see some RDF, FOAF and SemWeb apps... ----- Forwarded message from "Sean M. McNee" ----- From: "Sean M. McNee" Date: Tue, 21 Sep 2004 16:10:02 -0500 To: collab@sims.berkeley.edu Subject: [collab@sims] CFP: Beyond Personalization 2005 Message-Id: Call for Papers (with apologies for cross-posting) ------------------------------------------------------------ Beyond Personalization 2005 A Workshop on the Next Stage of Recommender Systems Research ------------------------------------------------------------ http://www.cs.umn.edu/Research/GroupLens/beyond2005/ ------------------------------------------------------------ In conjunction with the 2005 International Conference on Intelligent User Interfaces (IUI 2005) San Diego, California January 9-12, 2005 http://www.iuiconf.org ------------------------- Workshop Topics and Goals ------------------------- This workshop intends to bring recommender systems researchers and practitioners together in order to discuss the current state of recommender systems research, both on existing and emerging research topics, and to determine how research in this area should proceed. We are at a pivotal point in recommender systems research where researchers are both looking inward at what recommender systems are and looking outward at where recommender systems can be applied, and the implications of applying them out 'in the wild.' This creates a unique opportunity to both reassess the current state of research and directions research is taking in the near and long term. This workshop will focus on the following four main topics: 1. Understanding and trusting recommender systems. Do users understand and trust the recommendations they receive from recommender systems, what kinds of information do recommenders need to provide to users to build trust, and how difficult is it to regain trust in a recommender if it is lost? 2. User interfaces for recommender systems. What are good ways to present recommendations to users, how do you integrate recommenders into the displays of existing information systems, and how can interfaces encourage users to provide ratings in order to 'close the loop' for recommendations, that is, how can you get users to consume the items recommended and then tell the system how good the recommendations are? 3. The future of recommendation algorithms and their metrics. How can we generate better individual and group recommendations, develop new metrics and evaluation criteria for recommendations, and achieve cross-domain recommendations? 4. Social consequences and opportunities of recommender systems. How do individuals and groups of people respond to recommendations, how can recommendations be integrated with online and real world communities, and in what ways do recommendations affect social organizations? ----------------- Intended Audience ----------------- The workshop is intended for both established researchers and practioners in the domain of recommender systems as well as for new researchers and students with interesting ideas on recommender systems and their future. Participants do not have to come from a specific application domain, as long as their research or ideas are on one of the main topics of the workshop. ---------------- Paper Submission ---------------- Two types of contributions are invited: 1. Papers describing (ongoing) work on one or more of the topics for the workshop. 2. Position statements regarding one or more of the topics for the workshop. Both papers and position statements should be prepared according to the IUI 2005 Instructions for Authors, and should not exceed 6 pages for papers and 2 pages for position statements. Acceptable formats are PostScript and PDF. Please send the paper or position statement not later than 8 November 2004 by e-mail to mcnee@cs.umn.edu, cc: Mark.vanSetten@telin.nl Each paper and position statement will be reviewed by at least two reviewers. Accepted contributions will be published in the workshop proceedings and will be available on the Web before the workshop begins. Authors whose papers or position statements are accepted to this workshop are expected to attend both the workshop and IUI 2005. The workshop format will be panel-based with the authors of the best submitted papers appearing on the panels and include interactive sessions with all participants. --------------- Important Dates --------------- 8 Nov 2004: Paper Submission Deadline 28 Nov 2004: Paper Acceptance Notification 6 Dec 2004: Camera-ready Copies Due (and last day of Early Registration for IUI 2005) 10 Dec 2004: Papers Available from Website 9 Jan 2005: Workshop at IUI 2005 --------------- Workshop Chairs --------------- Mark van Setten Telematica Instituut P.O. Box 589 7500 AN Enschede The Netherlands E-mail: Mark.vanSetten@telin.nl Sean M. McNee GroupLens Research Department of Computer Science and Engineering University of Minnesota Minneapolis, MN, 55455 USA E-mail: mcnee@cs.umn.edu Joseph A. Konstan GroupLens Research Department of Computer Science and Engineering University of Minnesota Minneapolis, MN, 55455 USA E-mail: konstan@cs.umn.edu ----------------- Program Committee ----------------- Loren Terveen - University of Minnesota (USA) Liliana Ardissono - University of Torino (Italy) Jon Herlocker - Oregon State University (USA) Barry Smyth - University College Dublin and Changing Worlds (Ireland) Anton Nijholt - University of Twente (The Netherlands) --------- Posted to the collab@sims.berkeley.edu mailing list. To unsubscribe, send an email message to Majordomo@sims.berkeley.edu with the phrase "unsubscribe collab" in the body of the message. ----- End forwarded message ----- From danbri at w3.org Wed Sep 22 14:23:59 2004 From: danbri at w3.org (Dan Brickley) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] agenda item suggestion for wednesday's chat In-Reply-To: References: Message-ID: <20040922142359.GE5627@homer.w3.org> Hi Benjamin, all, * Benjamin Nowack [2004-09-21 20:36+0200] > > hi FOAFers, > > is the chat going to happen tomorrow (5:30 BST, IIRC)? Yes, though I didn't remember to announce it! Ahem... so that's in 3 hours time, for anyone who wants to stop by. #foaf IRC channel, irc.freenode.net aka irc://irc.freenode.net/foaf It should be pretty informal, given both the lack of notice and the fact I'm distracted with other things... > I'm currently trying to implement some tools related to > the FoafCommunityProcess[1]. If the chat is taking place, > I'd like to discuss some items of the "Things To Aspire > To" section: > > "identification of term maturity" > the spec currently uses "unstable", "testing", "stable" > which can be used to describe the current stage of a > term's life cycle. If a term is planned to be removed > from the term set we could use OWL's deprecation > constructs to flag a property or class as deprecated. > > I wonder if there could be (corner ?) cases where a > term's status went from "unstable" to "testing" but > after that didn't really get deployed. If the term > is not planned to be deprecated, is there a need/way > to indicate that this term "was tested, will probably > not move to stable but is going to be kept in the > spec". I think that's a case worth documenting in prose at least. Not clear how long it takes before we realise a term isn't going to end up in very widespread use. Also it might be argued that a few small-scale uses can still be of value. For eg., the stability vocab itself has some potential, but is unlikely to ever be used more than ~1000 or so times, because it is of fairly narrow applicability. > This is related to last week's "which terms > for names" discussion. Some terms might be in the > spec but their use is not recommended any more. Or > put in other words: "Tool developers, use these > terms. they'll move to 'stable' soon". perhaps this > discussion can be replaced by thinking about adding > "last modified" or "status changed" annotations to > each term. > There is some work in the Dublin Core Usage Board on this too. I'm hoping to attend their next f2f meeting as an observer, and have begun (phone chat last week) discussion with Tom Baker (chair) about potential for having a common model across both FOAF and DC. Tom is also leading the SW Best Practices taskforce on Vocabulary Management at W3C, so there's potential there I think. I'll try to find out what DC use, re last-modified etc. > "term documentation" > I volunteered at foafcamp to build some kind of a > "term browser" for easier access to the description > of selected foaf terms. Each term view could > show a term's standard information (label, comment, > term_status, relations, description, rdf/xml), but > also more detailed stuff such as examples, links to > tools, use cases, best practices, similar/related > terms, version info, issues, todos, etc. > Any thoughts on useful documentation types? Sounds cool! Translations too? :) > Just some ideas.. Thanks. See you later... Dan From Libby.Miller at bristol.ac.uk Wed Sep 22 18:44:54 2004 From: Libby.Miller at bristol.ac.uk (Libby Miller) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] foaf:dateOfBirth In-Reply-To: <20040921224230.GP8052@soto.kasei.com> References: <479B5AE2-E3C7-11D8-9F63-000D932A33F0@atommic.com> <20040921224230.GP8052@soto.kasei.com> Message-ID: hi Tony, hm, interesting point. At the moment you would have to talk about them separately, e.g. or somesuch. I'm not saying this is the right way mind you :) Plus it doesn't really say what you want....but it sort of provides a kind of evidence that we met, much like foaf:depicts does. RDF ical is at http://www.w3.org/2002/12/cal/ (currently slightly in flux) Libby On Tue, 21 Sep 2004, Tony Bowden wrote: > On Sun, Aug 01, 2004 at 03:46:31PM +0100, Libby Miller wrote: >> You could also use the bio vocab: >> [[ >> >> >> 1970-06-15 >> Brentwood, Essex, United Kingdom >> >> >> ]] >> http://vocab.org/bio/0.1/#Birth > > Are there other vocabs that include a wider range of events? Rather than > just saying that I foaf:knows someone, I'd often prefer to say that I > met them at a certain event. What's the preferred way to do something > like this? > > Tony > > > _______________________________________________ > rdfweb-dev mailing list > rdfweb-dev@vapours.rdfweb.org > wiki: http://rdfweb.org/topic/FoafProject > http://rdfweb.org/mailman/listinfo/rdfweb-dev > From bnowack at appmosphere.com Wed Sep 22 19:47:35 2004 From: bnowack at appmosphere.com (Benjamin Nowack) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] proposals for enhancing the descriptions of foaf terms Message-ID: [[ 17:04:57 action: bengee propose some clarifications re lifecycle/stability vocab to list, goal of having better machine-readable status for FOAF namespace ]] (from 2004-09-22's IRC chat [1]) There are actually two proposals we could discuss on this list (whether they make sense at all, how/if they could be implemented, etc.): 1) The foaf spec should give more info about a term's lifecycle stage than it is currently done via the "unstable"/"testing"/ "stable" term_status annotations. At the moment, it's not possible to see, e.g. *when* a term's status went to "testing". Or how long a term has been "unstable" (which could maybe tell a tool developer how actively it is maintained/how likely it is to move to "stable", etc). 2) the FOAF namespace should distinguish terms which "may be removed from spec", from terms whose "usage is discouraged" (because there are better idioms to use), but which will probably stay in the namespace indefinitely. some thoughts: re 1) We could use prose w/ owl:versionInfo, but having some kind of machine-readable "lastmodified" annotation would e.g. allow auto-generating an RSS feed for updated/added terms. This would also be possible if we used an agreed-on date/time format and owl:versionInfo. perhaps a DC term could be used. re 2) owl:Deprecated[Class|Property] could cover at least one of the cases. The owl reference doc says "by deprecating a term, it means that the term should not be used in new documents that commit to the ontology". As "may be removed from the spec" somehow includes "usage is discouraged", we possibly don't even need to distinguish the cases. A mentioned alternative would be the use of recommended term subsets (aka profiles ;) for different use cases or application areas. A re-worded proposal could then be "the foaf terms should have (machine-readable) pointers to application areas/use cases/implementing apps" which could allow the automatic generation of subsets for given use cases, or term sets of widely deployed terms." ideas, comments, objections? /action item bengee [1] http://www.ilrt.bris.ac.uk/discovery/chatlogs/rdfig/2004-09-22.html#T17-04-57 benjamin -- Benjamin Nowack Kruppstr. 100 45145 Essen, Germany http://www.appmosphere.com/ From bnowack at appmosphere.com Wed Sep 22 19:59:14 2004 From: bnowack at appmosphere.com (Benjamin Nowack) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] agenda item suggestion for wednesday's chat In-Reply-To: <20040922142359.GE5627@homer.w3.org> References: <20040922142359.GE5627@homer.w3.org> Message-ID: On 22.09.2004 10:23:59, Dan Brickley wrote: >"* Benjamin Nowack" <...> >> I volunteered at foafcamp to build some kind of a >> "term browser" for easier access to the description >> of selected foaf terms. Each term view could >> show a term's standard information (label, comment, >> term_status, relations, description, rdf/xml), but >> also more detailed stuff such as examples, links to >> tools, use cases, best practices, similar/related >> terms, version info, issues, todos, etc. >> Any thoughts on useful documentation types? > >... >Translations too? :) the tool supports different languages for class/property/ vocabulary annotations, and class/property doc items, but I'm not sure about character set limitations. At least chars from iso-8859-1 should work. benjamin >Thanks. See you later... >Dan > -- Benjamin Nowack Kruppstr. 100 45145 Essen, Germany http://www.appmosphere.com/ From tony-rdf at kasei.com Wed Sep 22 21:04:35 2004 From: tony-rdf at kasei.com (Tony Bowden) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] foaf:dateOfBirth In-Reply-To: References: <479B5AE2-E3C7-11D8-9F63-000D932A33F0@atommic.com> <20040921224230.GP8052@soto.kasei.com> Message-ID: <20040922210434.GA6109@soto.kasei.com> On Wed, Sep 22, 2004 at 07:44:54PM +0100, Libby Miller wrote: > hm, interesting point. At the moment you would have to talk about them > separately > > > > > > or somesuch. I'm not saying this is the right way mind you :) > Plus it doesn't really say what you want....but it sort of provides a > kind of evidence that we met, much like foaf:depicts does. I doesn't really provide much evidence though ... FOAFcamp was pretty small and I'm sure there were people there that I didn't meet. Start widening it out to something like OsCon, or even a sporting event, or concert or the like, and the chances that 2 people who were attendees actually met start getting much slimmer ... Co-depictions are slightly better, as there's a decent chance that 2 people in the same photograph (assuming it's of a small enough number of people) have indeed met [although I do have a bizarre story about how I once discovered than an ex-girlfriend had a photograph of the two of us from several years before we'd actually met ...] I'll certainly explore the ical stuff a bit more, though. Thanks, Tony Tony From mof-rdf at mfd-consult.dk Thu Sep 23 22:06:19 2004 From: mof-rdf at mfd-consult.dk (Morten Frederiksen) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] domain of foaf:interest Message-ID: <200409240006.19336.mof-rdf@mfd-consult.dk> Hi, Perhaps an agenda item for the next ScheduledTopicChat: When hacking up an "author list" for the WP FOAF plugin [1], I noticed that the blogroll of planetrdf [2] had some foaf:interest properties on the foaf:Document nodes describing the weblogs that are aggregated. This leads to a textbook contradiction [3], as the domain of foaf:interest is foaf:Person, and foaf:Document and foaf:Person are declared as owl:disjointWith each other. After chatting a bit with dajobe, who changed the blogroll right away, we realised that the change wasn't enough, as the blogroll contains weblogs of groups and projects, some of which may be said to have interests. However, because of the domain "contraint", these are now all person's, which is clearly wrong. One solution would be to remove the interest statements, perhaps using foaf:topic on the channels/items, but it doesn't seem to be all that irrelevant to talk of interests for at least groups, so I suggest relaxing the domain of foaf:interest to foaf:Agent instead of foaf:Person. [1] http://www.wasab.dk/morten/blog/archives/2004/07/05/wordpress-plugin-foaf-output [2] http://journal.dajobe.org/journal/2003/07/semblogs/bloggers.rdf [3] http://rdfweb.org/topic/FoafContradictions Regards, Morten From danbri at w3.org Fri Sep 24 00:19:12 2004 From: danbri at w3.org (Dan Brickley) Date: Tue Dec 12 14:55:26 2006 Subject: [rdfweb-dev] domain of foaf:interest In-Reply-To: <200409240006.19336.mof-rdf@mfd-consult.dk> References: <200409240006.19336.mof-rdf@mfd-consult.dk> Message-ID: <20040924001912.GF1975@homer.w3.org> * Morten Frederiksen [2004-09-24 00:06+0200] > Hi, > > Perhaps an agenda item for the next ScheduledTopicChat: > > When hacking up an "author list" for the WP FOAF plugin [1], I noticed that > the blogroll of planetrdf [2] had some foaf:interest properties on the > foaf:Document nodes describing the weblogs that are aggregated. > > This leads to a textbook contradiction [3], as the domain of foaf:interest is > foaf:Person, and foaf:Document and foaf:Person are declared as > owl:disjointWith each other. > > After chatting a bit with dajobe, who changed the blogroll right away, we > realised that the change wasn't enough, as the blogroll contains weblogs of > groups and projects, some of which may be said to have interests. However, > because of the domain "contraint", these are now all person's, which is > clearly wrong. > > One solution would be to remove the interest statements, perhaps using > foaf:topic on the channels/items, but it doesn't seem to be all that > irrelevant to talk of interests for at least groups, so I suggest relaxing > the domain of foaf:interest to foaf:Agent instead of foaf:Person. Good point. And who says OWL has only esoteric uses! It helps keep our modeling honest, at least... Anyone here object to such a change? If I don't here complaints, it'll happen. Dan > > > [1] > http://www.wasab.dk/morten/blog/archives/2004/07/05/wordpress-plugin-foaf-output > [2] http://journal.dajobe.org/journal/2003/07/semblogs/bloggers.rdf > [3] http://rdfweb.org/topic/FoafContradictions > > > Regards, > Morten > > _______________________________________________ > rdfweb-dev mailing list > rdfweb-dev@vapours.rdfweb.org > wiki: http://rdfweb.org/topic/FoafProject > http://rdfweb.org/mailman/listinfo/rdfweb-dev From james.carlyle at takepart.com Fri Sep 24 14:12:16 2004 From: james.carlyle at takepart.com (James Carlyle) Date: Tue Dec 12 14:55:27 2006 Subject: [rdfweb-dev] domain of foaf:interest In-Reply-To: <41542914.5050804@semanticplanet.com> References: <41542914.5050804@semanticplanet.com> Message-ID: <41542B40.5090207@takepart.com> Hi I would also be interested in discussing foaf:interest at the next ScheduledTopicChat. Ian Davis and I are building FriendSpace, trying to foster the concept "Community of Common Interests". We are bootstrapping off existing FOAF usage and the idea of "user-owned data", and are using foaf:interest to express voluntary membership of the group. We have 2 issues: 1) We want to present an aggregation of the following kind of personal activity for the community: Weblog posts, calendar events, bookmarks and possibly relationships (foaf:knows) Current foaf:interest usage describes the interests of a Person on a broad scale, but not the facets of a person's activities. So if we aggregate the daily activities of members of the group, we have no way of only looking for weblog posts or events that relate to a particular interest. In other words we need a way of finding how individual weblog posts, calendar events, bookmarks and foaf:knows relate to the foaf:interest that they say they have. There seem to be 2 alternatives here - see if people are willing to publish an interest-focused weblog, for example (or a category-focused sub-blog) i.e. Person interest Document (say uri is http://rdfweb.org/) made Channel page (domain is Resource, range is Document) Document (say uri is http://rdfweb.org/) made Channel (a.n.other) page ...a.n.other resource or to tag individual items with a "topic" i.e. Person interest Document (say uri is http://rdfweb.org/) seeAlso Vcalendar component Vevent page Document (say uri is http://rdfweb.org/) 2) The range of foaf:interest is foaf:Document, but it seems intuitive to say that I am interested in both tangible things, and intangible things that might not be reflected easily by "indicating a foaf:Document whose foaf:topic(s) broadly characterises that interest" (quoting from the FOAF spec). I am sure there is a lot of historical context here that I can't find, and would be interested in hearing the reasoning for this. I know Leigh, Morten and Kanzaki-san had views on the mailing list [1]. As you can see, we could tie a Vevent to a Document via foaf:page, but this seems an odd way of characterising RSS items or Vevents. Is there a better way? Kind regards James Carlyle [1] http://rdfweb.org/pipermail/rdfweb-dev/2004-March/012871.html >* Morten Frederiksen [2004-09-24 00:06+0200] > >>Hi, >> >>Perhaps an agenda item for the next ScheduledTopicChat: >> >>When hacking up an "author list" for the WP FOAF plugin [1], I noticed that >>the blogroll of planetrdf [2] had some foaf:interest properties on the >>foaf:Document nodes describing the weblogs that are aggregated. >> From mof-rdf at mfd-consult.dk Fri Sep 24 15:16:04 2004 From: mof-rdf at mfd-consult.dk (Morten Frederiksen) Date: Tue Dec 12 14:55:27 2006 Subject: [rdfweb-dev] domain of foaf:interest In-Reply-To: <41542914.5050804@semanticplanet.com> References: <41542914.5050804@semanticplanet.com> Message-ID: <200409241716.04096.mof-rdf@mfd-consult.dk> Hey, On Friday 24 September 2004 16:03, James Carlyle wrote: > 1) We want to present an aggregation of the following kind of personal > activity for the community: > Weblog posts, calendar events, bookmarks and possibly relationships > (foaf:knows) This is great, just the kind of tools we need. > There seem to be 2 alternatives here - see if people are willing to > publish an interest-focused weblog, for example (or a category-focused > sub-blog) I tried to do this with my FOAF output plugin for WP [1] (here I go, plugging it again), which also "enhances" the RSS feed [2] with topic information, like this: rss:item topic Document (X) At the same time, the FOAF file contains statements of interest: Person interest Document (X) The topic/interest information is based on categories and URIs assigned to them. > 2) The range of foaf:interest is foaf:Document, but it seems intuitive > to say that I am interested in both tangible things, and intangible > things that might not be reflected easily by "indicating a foaf:Document > whose foaf:topic(s) broadly characterises that interest" (quoting from > the FOAF spec). I am sure there is a lot of historical context here > that I can't find, and would be interested in hearing the reasoning for > this. I won't claim to understand the entire history, but one point is that foaf:topic and foaf:interest fit nicely together for exactly reasons like this, to be able to create connections. Remember that foaf:interest doesn't entail interest in the document, but in the *topic* of the document. Regards, Morten From danny.ayers at gmail.com Sat Sep 25 18:22:51 2004 From: danny.ayers at gmail.com (Danny Ayers) Date: Tue Dec 12 14:55:27 2006 Subject: [rdfweb-dev] foaf:dateOfBirth In-Reply-To: <20040921224230.GP8052@soto.kasei.com> References: <479B5AE2-E3C7-11D8-9F63-000D932A33F0@atommic.com> <20040921224230.GP8052@soto.kasei.com> Message-ID: <1f2ed5cd04092511223b30fcc5@mail.gmail.com> On Tue, 21 Sep 2004 23:42:30 +0100, Tony Bowden wrote: > Are there other vocabs that include a wider range of events? Rather than > just saying that I foaf:knows someone, I'd often prefer to say that I > met them at a certain event. What's the preferred way to do something > like this? I like the sound of what this suggests - xxx:met may be *too* simple, though it could be used with subclassing like foaf:knows. But given the existing vocabs for space/time etc I think it would probably be better /not/ to qualify it with where and when, etc. rather than having xxx:metAtBusStop What about a class something like Encounter, maybe (not quite the right properties, but you know what I mean): Encounter geo:location #place ical:when #date ical:occasion #event xxx:participant foaf:Person xxx:participant foaf:Person "participant" seems a bit clunky but I haven't the energy to click for a thesaurus. Cheers, Danny. From tony-rdf at kasei.com Sat Sep 25 19:18:20 2004 From: tony-rdf at kasei.com (Tony Bowden) Date: Tue Dec 12 14:55:27 2006 Subject: [rdfweb-dev] foaf:dateOfBirth In-Reply-To: <1f2ed5cd04092511223b30fcc5@mail.gmail.com> References: <479B5AE2-E3C7-11D8-9F63-000D932A33F0@atommic.com> <20040921224230.GP8052@soto.kasei.com> <1f2ed5cd04092511223b30fcc5@mail.gmail.com> Message-ID: <20040925191820.GA10992@soto.kasei.com> On Sat, Sep 25, 2004 at 08:22:51PM +0200, Danny Ayers wrote: > But given the existing vocabs for space/time etc I > think it would probably be better /not/ to qualify it with where and > when, etc. rather than having xxx:metAtBusStop I'm not sure I follow what this means ... > What about a class something like Encounter, maybe (not quite the > right properties, but you know what I mean): > Encounter > geo:location > #place > ical:when > #date > ical:occasion > #event > xxx:participant > foaf:Person > xxx:participant > foaf:Person At first glance looks good! Tony From james.carlyle at takepart.com Mon Sep 27 12:22:58 2004 From: james.carlyle at takepart.com (James Carlyle) Date: Tue Dec 12 14:55:27 2006 Subject: [rdfweb-dev] domain of foaf:interest In-Reply-To: <200409241716.04096.mof-rdf@mfd-consult.dk> References: <41542914.5050804@semanticplanet.com> <200409241716.04096.mof-rdf@mfd-consult.dk> Message-ID: <41580622.4080302@takepart.com> Morten Thanks for your encouragement and for your work on the WP Foaf plugin. Just a small point - shouldn't rss:item have a property called foaf:page instead of foaf:topic which points to a foaf:Document? From the Foaf spec, foaf:Document is in the range of foaf:interest and foaf:page, and the domain (not the range) of foaf:topic is foaf:Document. I am assuming you want to indicate how the interest of the foaf:Person relates to the rss:item. James > >I tried to do this with my FOAF output plugin for WP [1] (here I go, plugging >it again), which also "enhances" the RSS feed [2] with topic information, >like this: > >rss:item > topic > Document (X) > >At the same time, the FOAF file contains statements of interest: > >Person > interest > Document (X) > >The topic/interest information is based on categories and URIs assigned to >them. > > From iand at internetalchemy.org Mon Sep 27 13:26:45 2004 From: iand at internetalchemy.org (Ian Davis) Date: Tue Dec 12 14:55:27 2006 Subject: [rdfweb-dev] foaf:interest et al. Message-ID: <41581515.4060601@internetalchemy.org> Hi all, I've been having a think about foaf:interest and its close buddies foaf:topic and foaf:interest_topic which I've written about here[1] I think the foaf:interest property is misnamed. The problem is that the name is very general but the meaning is very specific (a page about a topic of interest) and somewhat at a tangent to the common meaning of the word. Not only is this confusing for authors it makes it harder to introduce new properties that are more general than foaf:interest. For example, I think the intended meaning of foaf:interest_topic (A thing of interest to this person.) is closer to the common meaning of interest. There's a lot of deployed FOAF out there that uses foaf:interest so this makes it very hard to change. However, it would be good to rationalise this part of the spec with a view to deprecating foaf:interest one day. My suggestion is to rename the foaf:topic_interest property to foaf:topicOfInterest or foaf:interestInTopic with the definition: A topic that this person expresses an interest in. Then, for completeness I'd also suggest the introduction of a foaf:Topic class. The domain of foaf:topicOfInterest and foaf:topic would be foaf:Topic. The range of foaf:topicOfInterest would be foaf:Agent. A better name for foaf:interest is harder, in my posting I suggest rather tongue in cheek foaf:interestInWhateverThisPageIsAbout. Maybe pageAboutInterest or interestPage or, to parallel a suggestion above, interestInPageTopic (i.e. bob has interest in this page's topic). Ian [1] http://www.semanticplanet.com/2004/09/topics.html From john.breslin at deri.org Mon Sep 27 14:28:37 2004 From: john.breslin at deri.org (John Breslin) Date: Tue Dec 12 14:55:27 2006 Subject: [rdfweb-dev] foaf:interest et al. In-Reply-To: <41581515.4060601@internetalchemy.org> Message-ID: <415831A5.16307.1820A02@localhost> If the foaf:topicOfInterest pointed to a category in a structure like the ODP, then maybe the problem of requiring a foaf:interestInWhateverThisPageIsAbout could be done away with :) I remember someone in FOAF Galway using foaf:interest with both a label and a URL - this seemed like good practice. I also wanted to use foaf:interest or foaf:topic_interest as part of my bulletin board FOAF export; people can define a like or dislike as shown here: http://www.boards.jp/network/likedis.php?do=view&u=1 where a label and URL to an ODP category are required, and another non-ODP URL is optional. But this leads me to ask how it would be possible to define a negative interest in something. This could be handy if for example I said that I had an interest in Sports, but do not have an interest in Baseball - so if things are recommended to me in a Sports domain, leave out the Baseball ones. foaf:topicOfNoInterest anyone? John. -- Dr. John Breslin Digital Enterprise Research Institute http://www.johnbreslin.com/ john.breslin@deri.org From iand at internetalchemy.org Mon Sep 27 14:50:18 2004 From: iand at internetalchemy.org (Ian Davis) Date: Tue Dec 12 14:55:27 2006 Subject: [rdfweb-dev] foaf:interest et al. In-Reply-To: <415831A5.16307.1820A02@localhost> References: <415831A5.16307.1820A02@localhost> Message-ID: <415828AA.8020604@internetalchemy.org> On 27/09/2004 15:28, John Breslin wrote: > > I remember someone in FOAF Galway using foaf:interest with both a label and a > URL - this seemed like good practice. This is good practice, but the label would apply to the page not to the topic. You could assign a label to the topic like this: The topic label > > But this leads me to ask how it would be possible to define a negative interest > in something. This could be handy if for example I said that I had an interest > in Sports, but do not have an interest in Baseball - so if things are > recommended to me in a Sports domain, leave out the Baseball ones. > foaf:topicOfNoInterest anyone? It should be possible to use Owl to do this by defining a class that is the the intersection of Sport with the complement of Baseball. Defining those two classes is left as an exercise for the reader... Ian From ndw at nwalsh.com Mon Sep 27 14:58:19 2004 From: ndw at nwalsh.com (Norman Walsh) Date: Tue Dec 12 14:55:27 2006 Subject: [rdfweb-dev] FOAF infrastructure changes requested Message-ID: <87vfdzkbck.fsf@nwalsh.com> Hello world, While working to improve the metadata presentation on norman.walsh.name, I stumbled over two issues. 1. The FOAF RDF Schema and the FOAF spec are content-negotiated from the namespace URI (http://xmlns.com/foaf/0.1/). This means that a tool like XSLT's document() function cannot get the schema from the namespace URI. I humbly suggest that the namespace URI should return the schema with a stylesheet PI that does something reasonable and includes a pointer to the human-readable spec. 2. The spec begins: ... I humbly request that it use one-fewer RDF serialization trick :-) and instead be coded as: Friend of a Friend (FOAF) vocabulary The Friend of a Friend (FOAF) RDF vocabulary, described using W3C RDF Schema and the Web Ontology Language. $Date: 2004/09/01 15:37:56 $ ... Granted, I'm asking because it makes my life easier, but I figure it can't hurt to ask :-) Be seeing you, norm -- Norman Walsh | Space isn't remote at all. It's only an http://nwalsh.com/ | hour's drive away if your car could go | straight upwards.--Fred Hoyle -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 188 bytes Desc: not available Url : http://lists.usefulinc.com/cgi-bin/mailman/private/foaf-dev/attachments/20040927/f7ebdc3b/attachment.pgp From iand at internetalchemy.org Mon Sep 27 15:09:03 2004 From: iand at internetalchemy.org (Ian Davis) Date: Tue Dec 12 14:55:27 2006 Subject: [rdfweb-dev] FOAF infrastructure changes requested In-Reply-To: <87vfdzkbck.fsf@nwalsh.com> References: <87vfdzkbck.fsf@nwalsh.com> Message-ID: <41582D0F.7020203@internetalchemy.org> Hi Norman, On 27/09/2004 15:58, Norman Walsh wrote: > While working to improve the metadata presentation on norman.walsh.name, > I stumbled over two issues. > > 1. The FOAF RDF Schema and the FOAF spec are content-negotiated from > the namespace URI (http://xmlns.com/foaf/0.1/). This means that a tool > like XSLT's document() function cannot get the schema from the > namespace URI. > > I humbly suggest that the namespace URI should return the schema with > a stylesheet PI that does something reasonable and includes a pointer > to the human-readable spec. Can I suggest an alternative? If you're dispatching based on namespace, perhaps you could maintain a simple namespace/schema mapping file, e.g.: You might be able to do something like: document( document( mappings.rdf )/rdf:Description[ @rdf:about="..."]/rdfs:seeAlso[0]/@rdf:resource ) Ian From jim.ley at gmail.com Mon Sep 27 15:18:36 2004 From: jim.ley at gmail.com (Jim Ley) Date: Tue Dec 12 14:55:27 2006 Subject: [rdfweb-dev] FOAF infrastructure changes requested In-Reply-To: <87vfdzkbck.fsf@nwalsh.com> References: <87vfdzkbck.fsf@nwalsh.com> Message-ID: <851c8d3104092708186fb460ea@mail.gmail.com> On Mon, 27 Sep 2004 10:58:19 -0400, Norman Walsh wrote: > 1. The FOAF RDF Schema and the FOAF spec are content-negotiated from > the namespace URI (http://xmlns.com/foaf/0.1/). This means that a tool > like XSLT's document() function cannot get the schema from the > namespace URI. I don't quite see why this follows, surely an XSLT parser must send an accept header containing only XML mime-types as acceptable. whilst including application/rdf+xml may be too much to expect, ensuring that the foaf URI returns XML in response to application/xml aswell should be fine. > I humbly suggest that the namespace URI should return the schema with > a stylesheet PI that does something reasonable and includes a pointer > to the human-readable spec. I really would rather not have XSLT PI's in the foaf spec - to make my life easier of course - I also prefer the human readable HTML as the default format. Have the best practices people looked at this sort of stuff yet? Cheers, Jim. From ndw at nwalsh.com Mon Sep 27 15:26:27 2004 From: ndw at nwalsh.com (Norman Walsh) Date: Tue Dec 12 14:55:27 2006 Subject: [rdfweb-dev] Re: FOAF infrastructure changes requested In-Reply-To: <41582D0F.7020203@internetalchemy.org> References: <87vfdzkbck.fsf@nwalsh.com> <41582D0F.7020203@internetalchemy.org> Message-ID: <877jqfhgws.fsf@nwalsh.com> LyBJYW4gRGF2aXMgPGlhbmRAaW50ZXJuZXRhbGNoZW15Lm9yZz4gd2FzIGhlYXJkIHRvIHNheToK fCBIaSBOb3JtYW4sCnwKfCBPbiAyNy8wOS8yMDA0IDE1OjU4LCBOb3JtYW4gV2Fsc2ggd3JvdGU6 Cnw+IFdoaWxlIHdvcmtpbmcgdG8gaW1wcm92ZSB0aGUgbWV0YWRhdGEgcHJlc2VudGF0aW9uIG9u IG5vcm1hbi53YWxzaC5uYW1lLAp8PiBJIHN0dW1ibGVkIG92ZXIgdHdvIGlzc3Vlcy4KfD4gMS4g VGhlIEZPQUYgUkRGIFNjaGVtYSBhbmQgdGhlIEZPQUYgc3BlYyBhcmUgY29udGVudC1uZWdvdGlh dGVkIGZyb20KfD4gICAgdGhlIG5hbWVzcGFjZSBVUkkgKGh0dHA6Ly94bWxucy5jb20vZm9hZi8w LjEvKS4gVGhpcyBtZWFucyB0aGF0IGEgdG9vbAp8PiAgICBsaWtlIFhTTFQncyBkb2N1bWVudCgp IGZ1bmN0aW9uIGNhbm5vdCBnZXQgdGhlIHNjaGVtYSBmcm9tIHRoZQp8PiAgICBuYW1lc3BhY2Ug VVJJLgp8PiAgICBJIGh1bWJseSBzdWdnZXN0IHRoYXQgdGhlIG5hbWVzcGFjZSBVUkkgc2hvdWxk IHJldHVybiB0aGUgc2NoZW1hCnw+IHdpdGgKfD4gICAgYSBzdHlsZXNoZWV0IFBJIHRoYXQgZG9l cyBzb21ldGhpbmcgcmVhc29uYWJsZSBhbmQgaW5jbHVkZXMgYSBwb2ludGVyCnw+ICAgIHRvIHRo ZSBodW1hbi1yZWFkYWJsZSBzcGVjLgp8CnwgQ2FuIEkgc3VnZ2VzdCBhbiBhbHRlcm5hdGl2ZT8g SWYgeW91J3JlIGRpc3BhdGNoaW5nIGJhc2VkIG9uCnwgbmFtZXNwYWNlLCBwZXJoYXBzIHlvdSBj b3VsZCBtYWludGFpbiBhIHNpbXBsZSBuYW1lc3BhY2Uvc2NoZW1hCnwgbWFwcGluZyBmaWxlLCBl LmcuOgp8CnwgPHJkZjpEZXNjcmlwdGlvbiByZGY6YWJvdXQ9Imh0dHA6Ly94bWxucy5jb20vZm9h Zi8wLjEvIj4KfCAgICA8cmRmczpzZWVBbHNvIHJkZjpyZXNvdXJjZT0iaHR0cDovL3htbG5zLmNv bS9mb2FmLzAuMS9pbmRleC5yZGYiLz4KfCA8L3JkZjpEZXNjcmlwdGlvbj4KfAp8IFlvdSBtaWdo dCBiZSBhYmxlIHRvIGRvIHNvbWV0aGluZyBsaWtlOgp8CnwgZG9jdW1lbnQoIGRvY3VtZW50KCBt YXBwaW5ncy5yZGYgKS9yZGY6RGVzY3JpcHRpb25bCnwgQHJkZjphYm91dD0iLi4uIl0vcmRmczpz ZWVBbHNvWzBdL0ByZGY6cmVzb3VyY2UgKQoKSSBjb3VsZCBkbyB0aGF0LCBidXQgSSBzdGlsbCBw cmVmZXIgbXkgYWx0ZXJuYXRpdmUuCgogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAg ICAgICAgQmUgc2VlaW5nIHlvdSwKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAg ICAgICAgbm9ybQoKLS0gCk5vcm1hbiBXYWxzaCA8bmR3QG53YWxzaC5jb20+IHwgQWxsIG91ciBm b2VzIGFyZSBtb3J0YWwuLS1QYXVsIFZhbMOpcnkKaHR0cDovL253YWxzaC5jb20vICAgICAgICAg ICAgfCAKLS0tLS0tLS0tLS0tLS0gbmV4dCBwYXJ0IC0tLS0tLS0tLS0tLS0tCkEgbm9uLXRleHQg YXR0YWNobWVudCB3YXMgc2NydWJiZWQuLi4KTmFtZTogbm90IGF2YWlsYWJsZQpUeXBlOiBhcHBs aWNhdGlvbi9wZ3Atc2lnbmF0dXJlClNpemU6IDE4OCBieXRlcwpEZXNjOiBub3QgYXZhaWxhYmxl ClVybCA6IGh0dHA6Ly9saXN0cy51c2VmdWxpbmMuY29tL2NnaS1iaW4vbWFpbG1hbi9wcml2YXRl L2ZvYWYtZGV2L2F0dGFjaG1lbnRzLzIwMDQwOTI3L2UwYTMxMTY3L2F0dGFjaG1lbnQucGdwCg== From mof-rdf at mfd-consult.dk Mon Sep 27 15:58:20 2004 From: mof-rdf at mfd-consult.dk (Morten Frederiksen) Date: Tue Dec 12 14:55:27 2006 Subject: [rdfweb-dev] domain of foaf:interest In-Reply-To: <41580622.4080302@takepart.com> References: <41542914.5050804@semanticplanet.com> <200409241716.04096.mof-rdf@mfd-consult.dk> <41580622.4080302@takepart.com> Message-ID: <200409271758.20428.mof-rdf@mfd-consult.dk> Hi James, all, On Monday 27 September 2004 14:22, James Carlyle wrote: > Just a small point - shouldn't rss:item have a property called foaf:page > instead of foaf:topic which points to a foaf:Document? From the Foaf > spec, foaf:Document is in the range of foaf:interest and foaf:page, and > the domain (not the range) of foaf:topic is foaf:Document. Right you are, wrong I was... Fortunately, I misrepresented what I had actually done (well, I had another error in there, but that was pointed out to me by Ian today [1]), so what I actually have is this [2]: rss:item topic (Something) dc:title page Document (X) Which, even if not exactly what you propose, should be OK (see also Ian's excellent writeup on the matter), except perhaps for the dc:title property, which perhaps should be rdfs:label or even foaf:name, as use of the Dublin Core properties seems best suited for documents. Opinions welcome. Sorry for the confusion. [1] http://www.wasab.dk/morten/blog/archives/2004/05/20/improving-rss-output-from-wordpress#comment-179 [2] http://www.wasab.dk/morten/blog/feed/rdf Regards, Morten From gk at ninebynine.org Mon Sep 27 14:28:43 2004 From: gk at ninebynine.org (Graham Klyne) Date: Tue Dec 12 14:55:27 2006 Subject: [rdfweb-dev] "Agent Model Yields Leadership" Message-ID: <5.1.0.14.2.20040927152553.02f3c948@127.0.0.1> This turned up in the ACM Technews digest: http://www.acm.org/technews/articles/2004-6/0922w.html#item5 The referenced article seems to suggest some interesting possibilities for FOAF-like networks. There are also resonances here with some of the trust-related work I've read about. #g -- # "Agent Model Yields Leadership" Technology Research News (09/29/04); Patch, Kimberly Researchers at Los Alamos National Laboratory and two universities have developed a software model for studying economic markets, quantitative sociology, or optimizing communications among robot collectives. The model is based on the classic minority game, where multiple agents compete to be in the minority of each round of decision-making; by adding a limited social network between the agents, the researchers were able to create a leadership structure that ultimately led to smarter and more adaptive performance than classic models without social network influence. Large and complicated systems such as the stock market are difficult to model because of the number of independent agents and choices available. Computing all possible scenarios is impossible with today's technology, but the researchers' model uses quantitative representations for agent behavior. Los Alamos National Laboratory researcher Zoltan Toroczkai says real human agents actually make decisions inductively rather than through deductive reasoning, as is assumed in classic game theory models; this is because, as with the computer models, figuring out all the possibilities is simply too difficult. The social network influence links each agent to its nearest neighbors and has them rely on the most recently successful agents for advice. Interestingly, the model grew more volatile when denser connectivity was added, since some leader agents' opinions became too popular and destabilized the system. Eventually, Toroczkai says the software model could help arrays of robots operate in conjunction where no human control is possible, such as on Mars explorations, although this type of technology would not be ready for another 10 to 20 years. The research was funded by the National Science Foundation, the Department of Energy, the Research Corporation, and the Alfred P. Sloan Foundation. Click Here to View Full Article http://www.trnmag.com/Stories/2004/092204/Agent_model_yields_leadership_092204.html ------------ Graham Klyne For email: http://www.ninebynine.org/#Contact From ndw at nwalsh.com Mon Sep 27 17:35:02 2004 From: ndw at nwalsh.com (Norman Walsh) Date: Tue Dec 12 14:55:27 2006 Subject: [rdfweb-dev] Re: FOAF infrastructure changes requested In-Reply-To: <851c8d3104092708186fb460ea@mail.gmail.com> References: <87vfdzkbck.fsf@nwalsh.com> <851c8d3104092708186fb460ea@mail.gmail.com> Message-ID: <87is9z8vjt.fsf@nwalsh.com> / Jim Ley was heard to say: | On Mon, 27 Sep 2004 10:58:19 -0400, Norman Walsh wrote: |> 1. The FOAF RDF Schema and the FOAF spec are content-negotiated from |> the namespace URI (http://xmlns.com/foaf/0.1/). This means that a tool |> like XSLT's document() function cannot get the schema from the |> namespace URI. | | I don't quite see why this follows, surely an XSLT parser must send an | accept header containing only XML mime-types as acceptable. whilst | including application/rdf+xml may be too much to expect, ensuring that | the foaf URI returns XML in response to application/xml aswell should | be fine. Alas, XSLT provides no way to influence the accept headers sent by the implementation. I imagine that application/xml is sent, but the server considers the XHTML version a better match. |> I humbly suggest that the namespace URI should return the schema with |> a stylesheet PI that does something reasonable and includes a pointer |> to the human-readable spec. | | I really would rather not have XSLT PI's in the foaf spec - to make my | life easier of course - I also prefer the human readable HTML as the | default format. Have the best practices people looked at this sort of | stuff yet? Dunno. I'm not sure content negotiating RDF and human text is a good idea, but I seem to be in the minority. Be seeing you, norm -- Norman Walsh | The Future is something which everyone http://nwalsh.com/ | reaches at the rate of sixty minutes an | hour, whatever he does, whoever he | is.--C. S. Lewis -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 188 bytes Desc: not available Url : http://lists.usefulinc.com/cgi-bin/mailman/private/foaf-dev/attachments/20040927/d60d2483/attachment.pgp From iand at internetalchemy.org Mon Sep 27 18:14:45 2004 From: iand at internetalchemy.org (Ian Davis) Date: Tue Dec 12 14:55:27 2006 Subject: [rdfweb-dev] FOAF infrastructure changes requested In-Reply-To: <851c8d3104092708186fb460ea@mail.gmail.com> References: <87vfdzkbck.fsf@nwalsh.com> <851c8d3104092708186fb460ea@mail.gmail.com> Message-ID: <41585895.40505@internetalchemy.org> On 27/09/2004 16:18, Jim Ley wrote: > On Mon, 27 Sep 2004 10:58:19 -0400, Norman Walsh wrote: > >>1. The FOAF RDF Schema and the FOAF spec are content-negotiated from >> the namespace URI (http://xmlns.com/foaf/0.1/). This means that a tool >> like XSLT's document() function cannot get the schema from the >> namespace URI. > > > I don't quite see why this follows, surely an XSLT parser must send an > accept header containing only XML mime-types as acceptable. whilst > including application/rdf+xml may be too much to expect, ensuring that > the foaf URI returns XML in response to application/xml aswell should > be fine. This might be problematic since application/xhtml+xml could also be returned. Both are obviously acceptable to an XSLT parser but they have differing levels of usefulness :) Ian From danbri at w3.org Mon Sep 27 18:56:53 2004 From: danbri at w3.org (Dan Brickley) Date: Tue Dec 12 14:55:27 2006 Subject: [rdfweb-dev] FOAF infrastructure changes requested In-Reply-To: <87vfdzkbck.fsf@nwalsh.com> References: <87vfdzkbck.fsf@nwalsh.com> Message-ID: <20040927185653.GA15164@homer.w3.org> * Norman Walsh [2004-09-27 10:58-0400] > Hello world, > > While working to improve the metadata presentation on norman.walsh.name, > I stumbled over two issues. > > 1. The FOAF RDF Schema and the FOAF spec are content-negotiated from > the namespace URI (http://xmlns.com/foaf/0.1/). This means that a tool > like XSLT's document() function cannot get the schema from the > namespace URI. > > I humbly suggest that the namespace URI should return the schema with > a stylesheet PI that does something reasonable and includes a pointer > to the human-readable spec. Do you think XSLT is deployable for such things yet? CSS styling can't, AFAIK, generate hyperlinks, so would be a hypertextual dead-end. If/when XSLT is "out there" enough, and browsers default to running untrusted XSLTs, then it could be the way to go. > 2. The spec begins: > > dc:title="Friend of a Friend (FOAF) vocabulary" > dc:description="The Friend of a Friend (FOAF) RDF vocabulary, described using W3C RDF Schema and the Web Ontology Language." > dc:date="$Date: 2004/09/01 15:37:56 $"> > ... > > I humbly request that it use one-fewer RDF serialization trick :-) and The reason we have an attribute-centric serialization is so the RDF can be shoved inside XHTML and the content not spill out in downlevel browsers. For the pure RDF version we could publish an index.rdf which had an element-based encoding. If someone wants to write the XSLT that generates one from the other, I could build it into the check scripts. > instead be coded as: > > > Friend of a Friend (FOAF) vocabulary > The Friend of a Friend (FOAF) RDF vocabulary, described using W3C RDF Schema and the Web Ontology Language. > $Date: 2004/09/01 15:37:56 $ > ... > > Granted, I'm asking because it makes my life easier, but I figure it can't > hurt to ask :-) FOAF is very much an environment where we can experiment with figuring out the options for what we rig up for de-referencable namespaces. I'm not sure there are many obvious right answers yet. Am happy to evolve things to try to meet as many people's needs as possible... cheers, Dan > _______________________________________________ > rdfweb-dev mailing list > rdfweb-dev@vapours.rdfweb.org > wiki: http://rdfweb.org/topic/FoafProject > http://rdfweb.org/mailman/listinfo/rdfweb-dev From danbri at w3.org Mon Sep 27 19:00:09 2004 From: danbri at w3.org (Dan Brickley) Date: Tue Dec 12 14:55:27 2006 Subject: [rdfweb-dev] Re: FOAF infrastructure changes requested In-Reply-To: <87is9z8vjt.fsf@nwalsh.com> References: <87vfdzkbck.fsf@nwalsh.com> <851c8d3104092708186fb460ea@mail.gmail.com> <87is9z8vjt.fsf@nwalsh.com> Message-ID: <20040927190009.GB15164@homer.w3.org> * Norman Walsh [2004-09-27 13:35-0400] > / Jim Ley was heard to say: > | On Mon, 27 Sep 2004 10:58:19 -0400, Norman Walsh wrote: > |> 1. The FOAF RDF Schema and the FOAF spec are content-negotiated from > |> the namespace URI (http://xmlns.com/foaf/0.1/). This means that a tool > |> like XSLT's document() function cannot get the schema from the > |> namespace URI. > | > | I don't quite see why this follows, surely an XSLT parser must send an > | accept header containing only XML mime-types as acceptable. whilst > | including application/rdf+xml may be too much to expect, ensuring that > | the foaf URI returns XML in response to application/xml aswell should > | be fine. > > Alas, XSLT provides no way to influence the accept headers sent by the > implementation. I imagine that application/xml is sent, but the server > considers the XHTML version a better match. Really? nothing in XSLT 2.0 or EXSLT? [time passes...] [asking Max Froumentin...] Nothing. Bummer. When's XSLT 3.0 planned? > > |> I humbly suggest that the namespace URI should return the schema with > |> a stylesheet PI that does something reasonable and includes a pointer > |> to the human-readable spec. > | > | I really would rather not have XSLT PI's in the foaf spec - to make my > | life easier of course - I also prefer the human readable HTML as the > | default format. Have the best practices people looked at this sort of > | stuff yet? > > Dunno. I'm not sure content negotiating RDF and human text is a good > idea, but I seem to be in the minority. It seemed like a good idea to me too, but then I figured it best to avoid # in namespace URIs for related reasons, and the jury still seems to be out on that one. Dan > _______________________________________________ > rdfweb-dev mailing list > rdfweb-dev@vapours.rdfweb.org > wiki: http://rdfweb.org/topic/FoafProject > http://rdfweb.org/mailman/listinfo/rdfweb-dev From jim.ley at gmail.com Mon Sep 27 19:01:37 2004 From: jim.ley at gmail.com (Jim Ley) Date: Tue Dec 12 14:55:27 2006 Subject: [rdfweb-dev] FOAF infrastructure changes requested In-Reply-To: <41585895.40505@internetalchemy.org> References: <87vfdzkbck.fsf@nwalsh.com> <851c8d3104092708186fb460ea@mail.gmail.com> <41585895.40505@internetalchemy.org> Message-ID: <851c8d310409271201232265b0@mail.gmail.com> On Mon, 27 Sep 2004 19:14:45 +0100, Ian Davis wrote: > This might be problematic since application/xhtml+xml could also be > returned. Both are obviously acceptable to an XSLT parser but they have > differing levels of usefulness :) ah, you also agree application/xhtml+xml is useless too, we just need to make sure danbri doesn't bother returning one of those then... :-) Jim. From iand at internetalchemy.org Mon Sep 27 22:13:49 2004 From: iand at internetalchemy.org (Ian Davis) Date: Tue Dec 12 14:55:27 2006 Subject: [rdfweb-dev] FOAF infrastructure changes requested In-Reply-To: <20040927185653.GA15164@homer.w3.org> References: <87vfdzkbck.fsf@nwalsh.com> <20040927185653.GA15164@homer.w3.org> Message-ID: <4158909D.6070400@internetalchemy.org> On 27/09/2004 19:56, Dan Brickley wrote: > The reason we have an attribute-centric serialization is so the RDF can > be shoved inside XHTML and the content not spill out in downlevel > browsers. For the pure RDF version we could publish an index.rdf which > had an element-based encoding. If someone wants to write the XSLT that > generates one from the other, I could build it into the check scripts. > Ah, I forgot the schema was buried in the xhtml. This should be readable by Norm's XSLT. (Is this the alternative to conneg? just pack it all inside one big XML document :) Ian From ndw at nwalsh.com Tue Sep 28 09:32:10 2004 From: ndw at nwalsh.com (Norman Walsh) Date: Tue Dec 12 14:55:27 2006 Subject: [rdfweb-dev] Re: FOAF infrastructure changes requested In-Reply-To: <20040927190009.GB15164@homer.w3.org> (Dan Brickley's message of "Mon, 27 Sep 2004 15:00:09 -0400") References: <87vfdzkbck.fsf@nwalsh.com> <851c8d3104092708186fb460ea@mail.gmail.com> <87is9z8vjt.fsf@nwalsh.com> <20040927190009.GB15164@homer.w3.org> Message-ID: <87oejq3fj9.fsf@nwalsh.com> / Dan Brickley was heard to say: | * Norman Walsh [2004-09-27 13:35-0400] |> / Jim Ley was heard to say: |> | On Mon, 27 Sep 2004 10:58:19 -0400, Norman Walsh wrote: |> |> 1. The FOAF RDF Schema and the FOAF spec are content-negotiated from |> |> the namespace URI (http://xmlns.com/foaf/0.1/). This means that a tool |> |> like XSLT's document() function cannot get the schema from the |> |> namespace URI. |> | |> | I don't quite see why this follows, surely an XSLT parser must send an |> | accept header containing only XML mime-types as acceptable. whilst |> | including application/rdf+xml may be too much to expect, ensuring that |> | the foaf URI returns XML in response to application/xml aswell should |> | be fine. |> |> Alas, XSLT provides no way to influence the accept headers sent by the |> implementation. I imagine that application/xml is sent, but the server |> considers the XHTML version a better match. | | Really? nothing in XSLT 2.0 or EXSLT? | [time passes...] [asking Max Froumentin...] EXSLT could do it. I suppose it would have made sense to do it for XSLT 2.0, but the document() function is already fairly complex. I think I'll drop it in as a feature request, but at this late stage, I expect to get slapped down pretty firmly. :-/ | It seemed like a good idea to me too, but then I figured it best to | avoid # in namespace URIs for related reasons, and the jury still seems | to be out on that one. Yep. Be seeing you, norm -- Norman Walsh | It is seldom that any liberty is lost http://nwalsh.com/ | all at once.--David Hume -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 188 bytes Desc: not available Url : http://lists.usefulinc.com/cgi-bin/mailman/private/foaf-dev/attachments/20040928/07a2515d/attachment.pgp From ndw at nwalsh.com Tue Sep 28 09:38:32 2004 From: ndw at nwalsh.com (Norman Walsh) Date: Tue Dec 12 14:55:27 2006 Subject: [rdfweb-dev] Re: FOAF infrastructure changes requested In-Reply-To: <20040927185653.GA15164@homer.w3.org> (Dan Brickley's message of "Mon, 27 Sep 2004 14:56:53 -0400") References: <87vfdzkbck.fsf@nwalsh.com> <20040927185653.GA15164@homer.w3.org> Message-ID: <87k6ue3f8n.fsf@nwalsh.com> / Dan Brickley was heard to say: | * Norman Walsh [2004-09-27 10:58-0400] |> Hello world, |> |> While working to improve the metadata presentation on norman.walsh.name, |> I stumbled over two issues. |> |> 1. The FOAF RDF Schema and the FOAF spec are content-negotiated from |> the namespace URI (http://xmlns.com/foaf/0.1/). This means that a tool |> like XSLT's document() function cannot get the schema from the |> namespace URI. |> |> I humbly suggest that the namespace URI should return the schema with |> a stylesheet PI that does something reasonable and includes a pointer |> to the human-readable spec. | | Do you think XSLT is deployable for such things yet? That's a judgement call, I guess. I publish random bits of XML with stylesheet PIs. Firefox and IE do the right thing. In the long run, I'd like to put a RDDL document at the namespace URI and let resolvers and other lower-level libraries "do the right thing". Of course, that means the APIs I'm using have to allow me to specify the media types I want back. Sigh. | CSS styling can't, | AFAIK, generate hyperlinks, so would be a hypertextual dead-end. Yeah, CSS doesn't work for these applications. | If/when | XSLT is "out there" enough, and browsers default to running untrusted | XSLTs, then it could be the way to go. I'm not sure I would advertise the namespace URI as the URI of the specification so I'd expect fewer people to go off dereferencing the namespace URI in their browsers. |> 2. The spec begins: |> |> dc:title="Friend of a Friend (FOAF) vocabulary" |> dc:description="The Friend of a Friend (FOAF) RDF vocabulary, described using W3C RDF Schema and the Web Ontology Language." |> dc:date="$Date: 2004/09/01 15:37:56 $"> |> ... |> |> I humbly request that it use one-fewer RDF serialization trick :-) and | | The reason we have an attribute-centric serialization is so the RDF can | be shoved inside XHTML and the content not spill out in downlevel | browsers. For the pure RDF version we could publish an index.rdf which | had an element-based encoding. If someone wants to write the XSLT that | generates one from the other, I could build it into the check scripts. Fair enough. A moot point since I can't get the RDF back from the server in my environment. I suppose if I setup a mapping table or something so that I can, I really ought to just load the URIs as RDF graphs anyway and not as XML documents, then it wouldn't matter. Be seeing you, norm -- Norman Walsh | So, are you working on finding that bug http://nwalsh.com/ | now, or are you leaving it until later? | Yes. -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 188 bytes Desc: not available Url : http://lists.usefulinc.com/cgi-bin/mailman/private/foaf-dev/attachments/20040928/f4b80ac5/attachment.pgp From ndw at nwalsh.com Tue Sep 28 10:03:47 2004 From: ndw at nwalsh.com (Norman Walsh) Date: Tue Dec 12 14:55:27 2006 Subject: [rdfweb-dev] Re: FOAF infrastructure changes requested In-Reply-To: <4158909D.6070400@internetalchemy.org> (Ian Davis's message of "Mon, 27 Sep 2004 23:13:49 +0100") References: <87vfdzkbck.fsf@nwalsh.com> <20040927185653.GA15164@homer.w3.org> <4158909D.6070400@internetalchemy.org> Message-ID: <87fz523e2k.fsf@nwalsh.com> / Ian Davis was heard to say: | On 27/09/2004 19:56, Dan Brickley wrote: | |> The reason we have an attribute-centric serialization is so the RDF can |> be shoved inside XHTML and the content not spill out in downlevel |> browsers. For the pure RDF version we could publish an index.rdf which |> had an element-based encoding. If someone wants to write the XSLT that |> generates one from the other, I could build it into the check scripts. |> | Ah, I forgot the schema was buried in the xhtml. This should be | readable by Norm's XSLT. Yeah, I suppose. I hadn't noticed it was stuck on the end. Be seeing you, norm -- Norman Walsh | If you are losing your leisure, look http://nwalsh.com/ | out! You may be losing your | soul.--Logan Pearsall Smith -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 188 bytes Desc: not available Url : http://lists.usefulinc.com/cgi-bin/mailman/private/foaf-dev/attachments/20040928/0647310d/attachment.pgp From Libby.Miller at bristol.ac.uk Tue Sep 28 19:10:33 2004 From: Libby.Miller at bristol.ac.uk (Libby Miller) Date: Tue Dec 12 14:55:27 2006 Subject: [rdfweb-dev] action: date of birth proposals Message-ID: ACTION libby: send mail to rdfweb-dev outlining 3 proposals for date or birth / birthday http://rdfig.xmlhack.com/2004/09/22/2004-09-22.html#1095870789.971234 Log: http://www.ilrt.bris.ac.uk/discovery/chatlogs/rdfig/2004-09-22.html#T17-32-02 Summary: (with snips) [[ 17:32:06 As I recall it...: 17:32:22 - a few old foaf files, like mine, have a plain literal field called dateOfBirth 17:32:28 that takes things like 1972-01-09 17:32:54 ...but it never got in spec, cos we also wanted to be able to say 'my birthday is jan 9th', without giving away age 17:33:11 useful for orkut-style birthday reminders, and a bit less revealing (some security issues re identity theft potential) ]] Three proposals: 1. Two new properties: foaf:dateOfBirth and foaf:birthday * dateOfBirth could be a string, datatype or object (non Gregorian calendars?), perhaps of the form 2004-09-22 (using http://www.w3.org/TR/xmlschema-2/#date) * similar considerations apply to birthday, except it would be of teh form 09-22 (MM-DD) (using http://www.w3.org/TR/xmlschema-2/#gMonthDay) 2. Three separate properties: dayOfBirth, monthOfBirth, yearOfBirth * similar considerations apply as above: http://www.w3.org/TR/xmlschema-2/#gYear http://www.w3.org/TR/xmlschema-2/#gMonth http://www.w3.org/TR/xmlschema-2/#gDay I'm not totally convinced the xml schema datatypes work quite correctly there, e.g. [[ gDay is a gregorian day that recurs, specifically a day of the month such as the 5th of the month ]] 3. Use existing bio:event structure bio vocab: http://vocab.org/bio/0.1/ - example on the front page: [[ 1970-06-15 Brentwood, Essex, United Kingdom ]] I'm not sure quite how birthday would work like this. Essentially it's a recurring event (which I tried to model in ical a while back: http://rdfweb.org/pipermail/rdfweb-dev/2003-November/012191.html) An additional idea might just be to use rdf ical: 4. Use RDF ical --the advantage here might be that you could download birthdays to PIMS. any thoughts/preferences? Libby From bkdelong at pobox.com Tue Sep 28 19:18:07 2004 From: bkdelong at pobox.com (B.K. DeLong) Date: Tue Dec 12 14:55:27 2006 Subject: [rdfweb-dev] action: date of birth proposals In-Reply-To: Message-ID: <5.2.1.1.2.20040928151429.02c8f080@mail.brain-stream.com> At 08:10 PM 9/28/2004 +0100, Libby Miller wrote: >2. Three separate properties: dayOfBirth, monthOfBirth, yearOfBirth > * similar considerations apply as above: >http://www.w3.org/TR/xmlschema-2/#gYear >http://www.w3.org/TR/xmlschema-2/#gMonth >http://www.w3.org/TR/xmlschema-2/#gDay > >I'm not totally convinced the xml schema datatypes work quite correctly >there, e.g. > >[[ >gDay is a gregorian day that recurs, specifically a day of the month such >as the 5th of the month >]] I think 3 is the way to go. As encryption and control of data becomes increasingly important in the FOAF community, people need the ability to be able to specify just how much of their "birthday equation" someone can have via trust-relationships. By separating it out into 3 different properties, an application could spit out a single property, a pair, or all three. >3. Use existing bio:event structure > >bio vocab: http://vocab.org/bio/0.1/ >- example on the front page: >[[ > > > > 1970-06-15 > Brentwood, Essex, United Kingdom > > > >]] > >I'm not sure quite how birthday would work like this. Essentially it's a >recurring event (which I tried to model in ical a while back: >http://rdfweb.org/pipermail/rdfweb-dev/2003-November/012191.html) If we use Bio, I'd like to see the date broken out into 3 parts. BTW, one thing we are missing - timeofBirth ;) It only really makes a real difference to twins and astrologers but it is a dataset we have nowadays. -- B.K. DeLong bkdelong@pobox.com +1.617.797.2472 http://bkdelong.mit.edu Work. http://www.brain-stream.com Play. http://www.the-leaky-cauldron.org Potter. http://www.city-of-doors.com Sigil. http://www.hackerfoundation.org Future. http://www.osvdb.org/ Security. http://www.wkdelong.org Son. PGP Fingerprint: 38D4 D4D4 5819 8667 DFD5 A62D AF61 15FF 297D 67FE FOAF: http://foaf.brain-stream.org From bkdelong at pobox.com Tue Sep 28 19:20:52 2004 From: bkdelong at pobox.com (B.K. DeLong) Date: Tue Dec 12 14:55:27 2006 Subject: [rdfweb-dev] action: date of birth proposals In-Reply-To: <5.2.1.1.2.20040928151429.02c8f080@mail.brain-stream.com> References: Message-ID: <5.2.1.1.2.20040928152028.02c825e8@mail.brain-stream.com> At 03:18 PM 9/28/2004 -0400, B.K. DeLong wrote: >I think 3 is the way to go. As encryption and control of data becomes >increasingly important in the FOAF community, people need the ability to >be able to specify just how much of their "birthday equation" someone can >have via trust-relationships. By separating it out into 3 different >properties, an application could spit out a single property, a pair, or >all three. Blast. I meant #2. -- B.K. DeLong bkdelong@pobox.com +1.617.797.2472 http://bkdelong.mit.edu Work. http://www.brain-stream.com Play. http://www.the-leaky-cauldron.org Potter. http://www.city-of-doors.com Sigil. http://www.hackerfoundation.org Future. http://www.osvdb.org/ Security. http://www.wkdelong.org Son. PGP Fingerprint: 38D4 D4D4 5819 8667 DFD5 A62D AF61 15FF 297D 67FE FOAF: http://foaf.brain-stream.org From jim at jibbering.com Tue Sep 28 19:30:23 2004 From: jim at jibbering.com (Jim Ley) Date: Tue Dec 12 14:55:27 2006 Subject: [rdfweb-dev] Re: action: date of birth proposals References: Message-ID: "Libby Miller" wrote in message news:Pine.GSO.4.61.0409281947450.6099@mail.ilrt.bris.ac.uk... > 2. Three separate properties: dayOfBirth, monthOfBirth, yearOfBirth > * similar considerations apply as above: > http://www.w3.org/TR/xmlschema-2/#gYear > http://www.w3.org/TR/xmlschema-2/#gMonth > http://www.w3.org/TR/xmlschema-2/#gDay I like this, it allows you to partial information, rather than being forced to give out everything. Jim. From Libby.Miller at bristol.ac.uk Tue Sep 28 19:38:15 2004 From: Libby.Miller at bristol.ac.uk (Libby Miller) Date: Tue Dec 12 14:55:27 2006 Subject: [rdfweb-dev] proposed agenda foaf meeting 2004-09-29, 1630 UTC Message-ID: hi all, The next Foaf project meeting is at 1630 UTC 2004-09-29 for 60-90 minutes. In your timezone: http://www.timeanddate.com/worldclock/fixedtime.html?day=29&month=9&year=2004&hour=16&min=30&sec=0&p1=0 I propose we use irc.freenode.net #rdfig again rather than #foaf (for the tools) Last meeting: Agenda, actions: http://rdfig.xmlhack.com/2004/09/22/2004-09-22.html logs: http://www.ilrt.bris.ac.uk/discovery/chatlogs/rdfig/2004-09-22.html#T16-28-52 Actions from last meeting: * danbri update FOAF spec to add a main section on naming, to provide a unified treatment of first/last/given/family addressing Misha's concern re first/last - continued * action: crschmidt to circulate crawler stats on foaf:nick foaf:name foaf:firstName foaf:givenname foaf:surname foaf:family_name - continued * ACTION libby: send mail to rdfweb-dev outlining 3 proposals for date or birth / birthday - see http://rdfweb.org/pipermail/rdfweb-dev/2004-September/013733.html * ACTION danbri: invite Peter St Andre to a future IRC FOAF meeting * ACTION danbri propose to the list "that it should be a goal to ultimately get the FOAF spec in a shape (whatever shape that is) such that IETF, Dublin Core and W3C Working Groups are comfortable citing it when they need to draw upon core concepts like Person, homepage, mbox, mbox_sha1sum etc." * action danbri: get his FOAF talks from FOAFCamp and Galway online (and do the workshop report!) * action: danbri let nicole know that Jim had problems with fontsize on new sit * ACTION bengee propose some clarifications re lifecycle/stability vocab to list, goal of having better machine-readable status for FOAF namespace - see http://rdfweb.org/pipermail/rdfweb-dev/2004-September/013705.html Agenda proposals for 2004-09-29 meeting * Domain of foaf:interest - Morten http://rdfweb.org/pipermail/rdfweb-dev/2004-September/013708.html + thread * Foaf views and profiles 17:32:27 pjz: views or profiles could probably be made another action/topic for the regular irc meetings? 17:34:10 * teefal seconds profiles/views as future topic http://www.ilrt.bris.ac.uk/discovery/chatlogs/rdfig/2004-09-22.html#T17-32-27 and possibly... * Norm's mail http://rdfweb.org/pipermail/rdfweb-dev/2004-September/013718.html More welcome! Libby From stpeter at jabber.org Tue Sep 28 20:29:03 2004 From: stpeter at jabber.org (Peter Saint-Andre) Date: Tue Dec 12 14:55:27 2006 Subject: [rdfweb-dev] Re: proposed agenda foaf meeting 2004-09-29, 1630 UTC References: Message-ID: In article , Libby Miller wrote: > * ACTION danbri: invite Peter St Andre to a future IRC FOAF meeting Wow, I have my very own action item! ;-) I'll make every attempt to be there tomorrow. /psa From stpeter at jabber.org Tue Sep 28 20:54:33 2004 From: stpeter at jabber.org (Peter Saint-Andre) Date: Tue Dec 12 14:55:27 2006 Subject: [rdfweb-dev] Re: action: date of birth proposals References: Message-ID: In article , Libby Miller wrote: > 2. Three separate properties: dayOfBirth, monthOfBirth, yearOfBirth > * similar considerations apply as above: > http://www.w3.org/TR/xmlschema-2/#gYear > http://www.w3.org/TR/xmlschema-2/#gMonth > http://www.w3.org/TR/xmlschema-2/#gDay This seems best (clearest and most flexible) to me as well. /psa From richard at cyganiak.de Tue Sep 28 22:10:00 2004 From: richard at cyganiak.de (Richard Cyganiak) Date: Tue Dec 12 14:55:27 2006 Subject: [rdfweb-dev] action: date of birth proposals In-Reply-To: References: Message-ID: <23DBFCAA-119B-11D9-99BF-000A95DABFDC@cyganiak.de> > ACTION libby: send mail to rdfweb-dev outlining 3 proposals for date > or birth / birthday ... > 1. Two new properties: foaf:dateOfBirth and foaf:birthday I like foaf:birthday because it lets you trivially find people with the same birthday (as in same month and day). But why not avoid the overlap? foaf:birthday, e.g. "07-15"^^xsd:gDayMonth foaf:yearOfBirth, e.g. "1979"^^xsd:gYear The year could still be omitted, and the reason for there being two properties would be clearer. Also smaller risk of accidently confusing the two props. I also like that both properties are useful in isolation. Birthday tells you when to congratulate. YearOfBirth tells you the age. Very natural. Splitting birthday into dayOfBirth and monthOfBirth, like in proposal 2, gives you two quite awkward properties that are only useful if both are known. Richard From danbri at w3.org Tue Sep 28 22:49:16 2004 From: danbri at w3.org (Dan Brickley) Date: Tue Dec 12 14:55:27 2006 Subject: [rdfweb-dev] Re: proposed agenda foaf meeting 2004-09-29, 1630 UTC In-Reply-To: References: Message-ID: <20040928224916.GF9755@homer.w3.org> * Peter Saint-Andre [2004-09-28 14:29-0600] > In article , > Libby Miller > wrote: > > > * ACTION danbri: invite Peter St Andre to a future IRC FOAF meeting > > Wow, I have my very own action item! ;-) > > I'll make every attempt to be there tomorrow. That'd be great. I declare my ACTION completed ;) Would you be able to lead 5-10 mins discussion on FOAF-related requirements/issues etc in a Jabber/XMPP context? You've previously mentioned concerns about imposing a requirement for an RDF parser on lightweight Jabber clients. I hope we can find a way to distinguish (i) having an RDF representation of Jabber profile data from (ii) having that encoded directly in RDF/XML, eg. http://www.w3.org/TR/2004/NOTE-grddl-20040413/ suggests one (XSLT-based) technique for mapping non-RDF/XML markup idioms into the RDF data model. If you could give us a sense of where things are up to in the Jabber world, and the kinds of profile information you'd be interested to carry, I think that'd be really helpful. Anything that you could send as email'd background reading would of course be much appreciated. cheers, Dan From stpeter at jabber.org Tue Sep 28 23:05:22 2004 From: stpeter at jabber.org (Peter Saint-Andre) Date: Tue Dec 12 14:55:27 2006 Subject: [rdfweb-dev] Re: proposed agenda foaf meeting 2004-09-29, 1630 UTC In-Reply-To: <20040928224916.GF9755@homer.w3.org> References: <20040928224916.GF9755@homer.w3.org> Message-ID: On Sep 28, 2004, at 4:49 PM, Dan Brickley wrote: > Would you be able to lead 5-10 mins discussion on FOAF-related > requirements/issues etc in a Jabber/XMPP context? I'd be happy to. > You've previously mentioned concerns about imposing a requirement for > an > RDF parser on lightweight Jabber clients. I hope we can find a way to > distinguish (i) having an RDF representation of Jabber profile data > from > (ii) having that encoded directly in RDF/XML, eg. > http://www.w3.org/TR/2004/NOTE-grddl-20040413/ suggests one > (XSLT-based) > technique for mapping non-RDF/XML markup idioms into the RDF data > model. There's been some discussion about this on the Standards-JIG list, which is where Jabber protocol talk happens (see this post and follow-ups if you're interested): http://mail.jabber.org/pipermail/standards-jig/2004-September/ 006254.html > If you could give us a sense of where things are up to in the Jabber > world, and the kinds of profile information you'd be interested to > carry, I think that'd be really helpful. Anything that you could send > as > email'd background reading would of course be much appreciated. Probably the best background info right now is in my blog: http://www.saint-andre.com/blog/2004-09.html#2004-09-17T09:25 I hope to write up something more formal (i.e., a JEP) in a week or two. Peter From am2stewa at uwaterloo.ca Wed Sep 29 02:32:30 2004 From: am2stewa at uwaterloo.ca (Alex Stewart) Date: Tue Dec 12 14:55:27 2006 Subject: [rdfweb-dev] Does rdfs:seeAlso imply anything? Message-ID: <20040929023230.GC334@scg.math> Hi, If we have blank1 rdfs:seeAlso uri1 and uri1 foaf:primaryTopic blank2 then can blank1 and blank2 be smushed into one node? I've searched the archive for "seeAlso topic" and I've found a message about PPDs from Morten[1] that suggests no, but then I've also seen the second mode of thought in PointingFromPersonToGroup in the wiki[2] that hints yes (since the foaf:Group doesn't have any inverse functional properties). Another yes vote comes from Leigh's Introduction to FOAF[3] (search for "Harry"; it's right after the first occurance). Mind you, people aren't going out and saying "yes" to this question, but rather it's just implied in the RDF statements that they write. Alex 1. http://rdfweb.org/pipermail/rdfweb-dev/2004-February/012663.html 2. http://rdfweb.org/topic/PointingFromPersonToGroup 3. http://www.xml.com/pub/a/2004/02/04/foaf.html From mof-rdf at mfd-consult.dk Wed Sep 29 06:06:47 2004 From: mof-rdf at mfd-consult.dk (Morten Frederiksen) Date: Tue Dec 12 14:55:27 2006 Subject: [rdfweb-dev] Does rdfs:seeAlso imply anything? In-Reply-To: <20040929023230.GC334@scg.math> References: <20040929023230.GC334@scg.math> Message-ID: <200409290806.47988.mof-rdf@mfd-consult.dk> Hi Alex, On Wednesday 29 September 2004 04:32, Alex Stewart wrote: > If we have > blank1 rdfs:seeAlso uri1 > and > uri1 foaf:primaryTopic blank2 > then can blank1 and blank2 be smushed into one node? I stand firmly by my view as you've pointed at, the answer is no. > I've searched the archive for "seeAlso topic" and I've found a message > about PPDs from Morten[1] that suggests no, but then I've also seen the > second mode of thought in PointingFromPersonToGroup in the wiki[2] that > hints yes (since the foaf:Group doesn't have any inverse functional > properties). I think you're reading too much into that. The connection between the group and the person in that example is formed by the use of the foaf:member property, not the rdfs:seeAlso, note the comment, "ME, IDENTIFIED IN THE NODEID ABOVE", which is hinting at something like: ... ... > Another yes vote comes from Leigh's Introduction to > FOAF[3] (search for "Harry"; it's right after the first occurance). Hmm, while I do see why it can be seen as a "yes vote", I think it's simply a case of the example missing identifying properties for all the person's mentioned. > Mind you, people aren't going out and saying "yes" to this question, but > rather it's just implied in the RDF statements that they write. Agreed, in many cases it is indeed true, but changing the semantics of rdfs:seeAlso (or overloading it) isn't needed, as in those cases some (other) identifying properties are usually used as well. Regards, Morten From mof-rdf at mfd-consult.dk Wed Sep 29 06:19:38 2004 From: mof-rdf at mfd-consult.dk (Morten Frederiksen) Date: Tue Dec 12 14:55:27 2006 Subject: [rdfweb-dev] foaf:interest et al. In-Reply-To: <41581515.4060601@internetalchemy.org> References: <41581515.4060601@internetalchemy.org> Message-ID: <200409290819.38559.mof-rdf@mfd-consult.dk> Hi, On Monday 27 September 2004 15:26, Ian Davis wrote: > I've been having a think about foaf:interest and its close buddies > foaf:topic and foaf:interest_topic which I've written about here[1] Very nice writeup. > I think the foaf:interest property is misnamed. I really don't disagree on that one, but the definition is solid and it's being used, which I think are two more important points. > My suggestion is to rename the foaf:topic_interest property to > foaf:topicOfInterest or foaf:interestInTopic with the definition: A > topic that this person expresses an interest in. Sounds quite reasonable, even if the topic is hard to define, see below. > Then, for completeness > I'd also suggest the introduction of a foaf:Topic class. The domain of > foaf:topicOfInterest and foaf:topic would be foaf:Topic. The range of > foaf:topicOfInterest would be foaf:Agent. Hmm, I think we should think (at least) twice about this, since you'll soon realise that practially anything can be a topic: _:me foaf:interest . If we agree that the above is a reasonable statement, then that'd make His Bobness a foaf:Topic, which might not be wrong, but seems somewhat strange. Exactly what is a "topic"? Perhaps some SKOS [1] stuff could be pulled in here? [1] http://www.w3.org/2001/sw/Europe/reports/thes/1.0/guide/ Regards, Morten From tony-rdf at kasei.com Wed Sep 29 08:26:54 2004 From: tony-rdf at kasei.com (Tony Bowden) Date: Tue Dec 12 14:55:27 2006 Subject: [rdfweb-dev] action: date of birth proposals In-Reply-To: <23DBFCAA-119B-11D9-99BF-000A95DABFDC@cyganiak.de> References: <23DBFCAA-119B-11D9-99BF-000A95DABFDC@cyganiak.de> Message-ID: <20040929082654.GA26944@soto.kasei.com> On Wed, Sep 29, 2004 at 12:10:00AM +0200, Richard Cyganiak wrote: > I also like that both properties are useful in isolation. Birthday > tells you when to congratulate. YearOfBirth tells you the age. Very > natural. Erm? You can't tell age just from year of birth ... Tony From danbri at w3.org Wed Sep 29 09:05:00 2004 From: danbri at w3.org (Dan Brickley) Date: Tue Dec 12 14:55:27 2006 Subject: [rdfweb-dev] Does rdfs:seeAlso imply anything? In-Reply-To: <20040929023230.GC334@scg.math> References: <20040929023230.GC334@scg.math> Message-ID: <20040929090500.GA6822@homer.w3.org> Hi Alex, * Alex Stewart [2004-09-28 22:32-0400] > Hi, > > If we have > > blank1 rdfs:seeAlso uri1 > and > uri1 foaf:primaryTopic blank2 > > then can blank1 and blank2 be smushed into one node? > > I've searched the archive for "seeAlso topic" and I've found a message > about PPDs from Morten[1] that suggests no, but then I've also seen the > second mode of thought in PointingFromPersonToGroup in the wiki[2] that > hints yes (since the foaf:Group doesn't have any inverse functional > properties). Another yes vote comes from Leigh's Introduction to > FOAF[3] (search for "Harry"; it's right after the first occurance). > > Mind you, people aren't going out and saying "yes" to this question, but > rather it's just implied in the RDF statements that they write. I think the answer here is 'no'; a resource can be the rdfs:seeAlso of something without necessarily being its foaf:primaryTopic. An example might help, danbri libby A document about libby that mentions danbri in passing blank1 is a node that stands for me. We are truly told that, w.r.t. me, we can seealso a.rdf for maybe some more information. This is pretty weak, weaker than either foaf:primaryTopic or foaf:topic. But it's true and useful. The example has a.rdf being mostly about someone else, with a passing mention of me (eg. a co-depiction, or co-authorship). In such cases it's often a judgement call whether to use foaf:topic vs foaf:primaryTopic. I've mentioned before here that deferring to the document's own claims about it's (primary)topic is probably a good strategy. http://esw.w3.org/topic/UsingSeeAlso has some more notes on rdfs:seeAlso. Hope this helps, Dan > Alex > > 1. http://rdfweb.org/pipermail/rdfweb-dev/2004-February/012663.html > 2. http://rdfweb.org/topic/PointingFromPersonToGroup > 3. http://www.xml.com/pub/a/2004/02/04/foaf.html > > _______________________________________________ > rdfweb-dev mailing list > rdfweb-dev@vapours.rdfweb.org > wiki: http://rdfweb.org/topic/FoafProject > http://rdfweb.org/mailman/listinfo/rdfweb-dev From GK at ninebynine.org Wed Sep 29 07:31:27 2004 From: GK at ninebynine.org (Graham Klyne) Date: Tue Dec 12 14:55:27 2006 Subject: [rdfweb-dev] action: date of birth proposals In-Reply-To: Message-ID: <5.1.0.14.2.20040929082357.00bbf4e0@127.0.0.1> Considering in the abstract (i.e. not considering practical details) I'd prefer something based on RDF iCal, on the basis that will (hopefully) help to encourage uniform representation (using RDF iCal) of dates across different RDF applications. The practical concern is that setting up an Ical recurring event structure for such a simple piece of information may be overly complicated for such a simple task, which suggests some kind of simple date-as-literal may be desired. I note that ISO 8601 (upon which I understand the XML schema date type, and RFC 3339, are based) allows a string of the form: --MMDD or --MM-DD for month+day only. (See ISO 8601:1988, section 5.2.1.3) #g -- At 20:10 28/09/04 +0100, Libby Miller wrote: >ACTION libby: send mail to rdfweb-dev outlining 3 proposals for date or >birth / birthday >http://rdfig.xmlhack.com/2004/09/22/2004-09-22.html#1095870789.971234 > >Log: >http://www.ilrt.bris.ac.uk/discovery/chatlogs/rdfig/2004-09-22.html#T17-32-02 > >Summary: (with snips) >[[ >17:32:06 As I recall it...: >17:32:22 - a few old foaf files, like mine, have a plain literal >field called dateOfBirth >17:32:28 that takes things like 1972-01-09 >17:32:54 ...but it never got in spec, cos we also wanted to be >able to say 'my birthday is jan 9th', without giving away age >17:33:11 useful for orkut-style birthday reminders, and a bit >less revealing (some security issues re identity theft potential) >]] > >Three proposals: > >1. Two new properties: foaf:dateOfBirth and foaf:birthday > * dateOfBirth could be a string, datatype or object (non Gregorian > calendars?), perhaps of the form 2004-09-22 (using > http://www.w3.org/TR/xmlschema-2/#date) > * similar considerations apply to birthday, except it would be of teh > form 09-22 (MM-DD) (using http://www.w3.org/TR/xmlschema-2/#gMonthDay) > >2. Three separate properties: dayOfBirth, monthOfBirth, yearOfBirth > * similar considerations apply as above: >http://www.w3.org/TR/xmlschema-2/#gYear >http://www.w3.org/TR/xmlschema-2/#gMonth >http://www.w3.org/TR/xmlschema-2/#gDay > >I'm not totally convinced the xml schema datatypes work quite correctly >there, e.g. > >[[ >gDay is a gregorian day that recurs, specifically a day of the month such >as the 5th of the month >]] > >3. Use existing bio:event structure > >bio vocab: http://vocab.org/bio/0.1/ >- example on the front page: >[[ > > > > 1970-06-15 > Brentwood, Essex, United Kingdom > > > >]] > >I'm not sure quite how birthday would work like this. Essentially it's a >recurring event (which I tried to model in ical a while back: >http://rdfweb.org/pipermail/rdfweb-dev/2003-November/012191.html) > >An additional idea might just be to use rdf ical: > >4. Use RDF ical > >--the advantage here might be that you could download birthdays to PIMS. > > >any thoughts/preferences? > >Libby > >_______________________________________________ >rdfweb-dev mailing list >rdfweb-dev@vapours.rdfweb.org >wiki: http://rdfweb.org/topic/FoafProject >http://rdfweb.org/mailman/listinfo/rdfweb-dev ------------ Graham Klyne For email: http://www.ninebynine.org/#Contact From bnowack at appmosphere.com Wed Sep 29 09:35:24 2004 From: bnowack at appmosphere.com (Benjamin Nowack) Date: Tue Dec 12 14:55:27 2006 Subject: [rdfweb-dev] action: date of birth proposals In-Reply-To: <20040929082654.GA26944@soto.kasei.com> References: <23DBFCAA-119B-11D9-99BF-000A95DABFDC@cyganiak.de> <20040929082654.GA26944@soto.kasei.com> Message-ID: On 29.09.2004 09:26:54, Tony Bowden wrote: >On Wed, Sep 29, 2004 at 12:10:00AM +0200, Richard Cyganiak wrote: >> I also like that both properties are useful in isolation. Birthday >> tells you when to congratulate. YearOfBirth tells you the age. Very >> natural. > >Erm? > >You can't tell age just from year of birth ... But I'd say you can tell it in a "sufficiently precise" way for the majority of related use cases. If I see someone with a yearOfBirth=1973, wouldn't that be sufficiently precise for saying that we have the same age, no matter if (s)he is still 30 or already 31? benjamin -- Benjamin Nowack Kruppstr. 100 45145 Essen, Germany http://www.appmosphere.com/ > >Tony > >_______________________________________________ >rdfweb-dev mailing list >rdfweb-dev@vapours.rdfweb.org >wiki: http://rdfweb.org/topic/FoafProject >http://rdfweb.org/mailman/listinfo/rdfweb-dev > > From bnowack at appmosphere.com Wed Sep 29 09:59:35 2004 From: bnowack at appmosphere.com (Benjamin Nowack) Date: Tue Dec 12 14:55:27 2006 Subject: [rdfweb-dev] action: date of birth proposals In-Reply-To: <5.1.0.14.2.20040929082357.00bbf4e0@127.0.0.1> References: <5.1.0.14.2.20040929082357.00bbf4e0@127.0.0.1> Message-ID: On 29.09.2004 08:31:27, Graham Klyne wrote: >Considering in the abstract (i.e. not considering practical details) I'd >prefer something based on RDF iCal, on the basis that will (hopefully) help >to encourage uniform representation (using RDF iCal) of dates across >different RDF applications. > >The practical concern is that setting up an Ical recurring event structure >for such a simple piece of information may be overly complicated for such a >simple task, which suggests some kind of simple date-as-literal may be >desired. I note that ISO 8601 (upon which I understand the XML schema date >type, and RFC 3339, are based) allows a string of the form: > --MMDD >or > --MM-DD >for month+day only. (See ISO 8601:1988, section 5.2.1.3) there were two ideas behind proposing three separate props. one is the flexibility of what people might want to say about themselves, which could be solved by a single (e.g. --mm-dd / yyyy-mm-dd) or two properties (birthday, yearOfBirth) as people are probably not going to tell their dayOfBirth without the monthOfBirth. but the other objective was the ease of querying simple triple stores. (or even standard ones as long as we can't be sure that regexps/date handling will be part of the DAWG recommendation). having separate props should make it possible to query even the simplest triple table for e.g.: - people who have their birthday this/next month (via the month prop) - people with a birthday for a given month-day-combination (via day/month) - people of a certain age (via year) but I admit, this is concerning only a limited practical view. RDF iCal compatibility could be another (maybe more) important one. benjamin -- Benjamin Nowack Kruppstr. 100 45145 Essen, Germany http://www.appmosphere.com/ > >#g >-- > >------------ >Graham Klyne >For email: >http://www.ninebynine.org/#Contact From ldodds at ingenta.com Wed Sep 29 12:41:00 2004 From: ldodds at ingenta.com (Leigh Dodds) Date: Tue Dec 12 14:55:27 2006 Subject: [rdfweb-dev] Does rdfs:seeAlso imply anything? In-Reply-To: <200409290806.47988.mof-rdf@mfd-consult.dk> References: <20040929023230.GC334@scg.math> <200409290806.47988.mof-rdf@mfd-consult.dk> Message-ID: <415AAD5C.1020902@ingenta.com> Morten Frederiksen wrote: >>Another yes vote comes from Leigh's Introduction to >>FOAF[3] (search for "Harry"; it's right after the first occurance). > > Hmm, while I do see why it can be seen as a "yes vote", I think it's simply a > case of the example missing identifying properties for all the person's > mentioned. Yes, it shouldn't be taken as a "yes vote", it was just meant to be a concise example showing use of foaf:knows and rdfs:seeAlso linking, you shouldn't read anything else into it. L. From iand at internetalchemy.org Wed Sep 29 12:43:20 2004 From: iand at internetalchemy.org (Ian Davis) Date: Tue Dec 12 14:55:27 2006 Subject: [rdfweb-dev] foaf:interest et al. In-Reply-To: <200409290819.38559.mof-rdf@mfd-consult.dk> References: <41581515.4060601@internetalchemy.org> <200409290819.38559.mof-rdf@mfd-consult.dk> Message-ID: <415AADE8.4090809@internetalchemy.org> On 29/09/2004 07:19, Morten Frederiksen wrote: >>I've been having a think about foaf:interest and its close buddies >>foaf:topic and foaf:interest_topic which I've written about here[1] > > Very nice writeup. Thank you >>I think the foaf:interest property is misnamed. > > I really don't disagree on that one, but the definition is solid and it's > being used, which I think are two more important points. I accept both those points but I think there is a fundamental issue at stake here. Why do we use human readable terms in data designed foer machine reading? The answer, of course, is that being human readable has many advantages in terms of authoring and validation. The names we give to vocabulary terms are important. They serve as mnemonics so we don't have to refer to the specification whenever we write some RDF. They also draw on shared context which makes them easier to use and easier to understand. The problem I see with foaf:interest is that the name given to it has a very general pre-existing meaning, whereas the FOAF definition is very specific. This is in contrast to foaf:knows which is a general term with a general definition, or foaf:jabberID with is very specific. >>Then, for completeness >>I'd also suggest the introduction of a foaf:Topic class. The domain of >>foaf:topicOfInterest and foaf:topic would be foaf:Topic. The range of >>foaf:topicOfInterest would be foaf:Agent. I completely messed this up. I meant: The range of foaf:topicOfInterest and foaf:topic would be foaf:Topic The domain of foaf:topicOfInterest would be foaf:Agent In that: [a foaf:Agent] foaf:topicOfInterest [a foaf:Topic] . > > Hmm, I think we should think (at least) twice about this, since you'll soon > realise that practially anything can be a topic: > > _:me foaf:interest . > > If we agree that the above is a reasonable statement, then that'd make His > Bobness a foaf:Topic, which might not be wrong, but seems somewhat strange. I think it's quite appropriate to consider Bob Dylan as a person a topic of interest. His music, poetry, lifestyle, philosophy, smoking habits and fan demographics are also valid topics. > Exactly what is a "topic"? I can be flippant and say it's the range of foaf:topic. More seriously, WordNet sense 2 is closest "some situation or event that is thought about" http://www.cogsci.princeton.edu/cgi-bin/webwn2.0?stage=1&word=topic > Perhaps some SKOS [1] stuff could be pulled in here? Probably. I need to read it instead of skim it. Ian From iand at internetalchemy.org Wed Sep 29 12:48:37 2004 From: iand at internetalchemy.org (Ian Davis) Date: Tue Dec 12 14:55:27 2006 Subject: [rdfweb-dev] proposed agenda foaf meeting 2004-09-29, 1630 UTC In-Reply-To: References: Message-ID: <415AAF25.8010405@internetalchemy.org> I fully intended to make this meeting but can't because of a family situation. I'm feeling bad because it was me who specifically requested a late afternoon meeting at Galway and I haven't been able to attend one yet! Please accept my apologies and hopefully I'll make the next one. Ian From danbri at w3.org Wed Sep 29 14:12:00 2004 From: danbri at w3.org (Dan Brickley) Date: Tue Dec 12 14:55:27 2006 Subject: [rdfweb-dev] a FOAF goal: spec fit for citation from standards docs Message-ID: <20040929141158.GI2012@homer.w3.org> In last IRC meeting, http://rdfig.xmlhack.com/2004/09/22/2004-09-22.html I said I'd drop the FOAF list a note proposing... "that it should be a goal to ultimately get the FOAF spec in a shape (whatever shape that is) such that IETF, Dublin Core and W3C Working Groups are comfortable citing it when they need to draw upon core concepts like Person, homepage, mbox, mbox_sha1sum etc." This is pretty broad-brush, necessarily, since these various organizations have different --- and evolving --- processes and conventions for citing externally managed specs. Without getting into distinction between normative and informational citations, I'd like to get a sense of whether folk here think this is generally a healthy goal. I do. It seems a shame to do all this work, only to have standards-track efforts feel that they'd need to duplicate it in a more formal setting. At least if bits of it do get duplicated/shadowed/etc., we should, I think, get FOAF into a form such that the editors of these specs feel comfortable making prose and RDF/XML references to the terms defined in the FOAF spec. Guess that means I need to take out the nudie pics from the spec... Dan From danbri at w3.org Wed Sep 29 14:13:08 2004 From: danbri at w3.org (Dan Brickley) Date: Tue Dec 12 14:55:27 2006 Subject: [rdfweb-dev] proposed agenda foaf meeting 2004-09-29, 1630 UTC In-Reply-To: References: Message-ID: <20040929141308.GJ2012@homer.w3.org> * Libby Miller [2004-09-28 20:38+0100] > > hi all, > > * ACTION danbri: invite Peter St Andre to a future IRC FOAF meeting Done; welcome, peter. > * ACTION danbri propose to the list "that it should be a goal to > ultimately get the FOAF spec in a shape (whatever shape that is) such > that IETF, Dublin Core and W3C Working Groups are comfortable citing it > when they need to draw upon core concepts like Person, homepage, mbox, > mbox_sha1sum etc." Done > * action danbri: get his FOAF talks from FOAFCamp and Galway online (and > do the workshop report!) In progress (aka not done); continued. > * action: danbri let nicole know that Jim had problems with fontsize on > new sit Done; From iand at internetalchemy.org Wed Sep 29 16:07:53 2004 From: iand at internetalchemy.org (Ian Davis) Date: Tue Dec 12 14:55:27 2006 Subject: [rdfweb-dev] a FOAF goal: spec fit for citation from standards docs In-Reply-To: <20040929141158.GI2012@homer.w3.org> References: <20040929141158.GI2012@homer.w3.org> Message-ID: <415ADDD9.4030704@internetalchemy.org> On 29/09/2004 15:12, Dan Brickley wrote: > In last IRC meeting, http://rdfig.xmlhack.com/2004/09/22/2004-09-22.html > I said I'd drop the FOAF list a note proposing... > > "that it should be a goal to ultimately get the FOAF spec in a shape > (whatever shape that is) such that IETF, Dublin Core and W3C Working Groups > are comfortable citing it when they need to draw upon core concepts > like Person, homepage, mbox, mbox_sha1sum etc." > > ... I'd like to > get a sense of whether folk here think this is generally a healthy goal. I also think this is a goal worth striving for. It doesn't have to mean a full-on standards process, but it does mean having a well-defined change process (which I think we are close to having). It also means we need to make certain guarantees about stability (again, this is being worked on). Being slightly tangential for a moment, I just want to re-float my suggestion that new terms be introduced in a separate namespace until they are mature enough to move into the core. I suggested this at Galway although I'm not sure I articulated it very clearly. My proposal is to create a new namespace http://xmlns.com/foaf/experimental/ into which new experimental terms are put. Implementors are made aware that terms in this namespace may change meaning, subdivide or disappear completely with little or no notice. Once a term has been around for long enough for us to be confident we understand how it fits into FOAF and its usefulness then it can migrate into the core namespace. The key advantage of this approach is that the creativity of the FOAF community isn't stifled by the requirements to be stable and serious about the spec. There are a number of disadvantages as you pointed out in Galway, such applications relying on particular namespace strings. Ian From Libby.Miller at bristol.ac.uk Thu Sep 30 13:49:12 2004 From: Libby.Miller at bristol.ac.uk (Libby Miller) Date: Tue Dec 12 14:55:27 2006 Subject: [rdfweb-dev] next foaf meeting? Message-ID: Dan's not around next wednesday, but I'm happy to hold a foaf meeting if there's demand. It would be 1630 UTC, 6th October, on #rdfig. Ping me offlist if you think you can make it and I'll send something in the next couple of days yes or no. Agenda items welcome too, preferably to the list. In your timezone: http://www.timeanddate.com/worldclock/fixedtime.html?day=6&month=10&year=2004&hour=16&min=30&sec=0&p1=0 cheers, Libby From danbri at w3.org Wed Sep 1 16:13:28 2004 From: danbri at w3.org (Dan Brickley) Date: Mon Dec 18 17:23:56 2006 Subject: [rdfweb-dev] stuff to look at Message-ID: <20040901161328.GB17604@homer.w3.org> some urls that have come up in conversation at FOAF Galway: http://web.mit.edu/ryanlee/www/thesis/thesis.html Personal Data Protection in the Semantic Web Masters of Engineering Thesis http://www.w3.org/TR/rdf-dawg-uc/ RDF Data Access Use Cases and Requirements http://www.w3.org/TR/2004/WD-swbp-n-aryRelations-20040721/ Defining N-ary Relations on the Semantic Web: Use With Individuals SKOS: http://www.w3.org/2004/02/skos/core/ RDFAuthor's RDF querying tutorial, http://rdfweb.org/people/damian/2001/10/RDFAuthor/Tutorial/Tutorial2/ From ronwalf at volus.net Thu Sep 2 16:10:33 2004 From: ronwalf at volus.net (Ron Alford) Date: Mon Dec 18 17:23:56 2006 Subject: [rdfweb-dev] Using OWL instead of foaf:membershipClass property Message-ID: <413745F9.6070403@volus.net> I haven't seen much response to my suggestions to make foaf owl-friendlier, so I'm putting this forward both as to help make foaf semantics more machine readable, and to motivate some clean ups to make foaf friendlier to owl reasoners. IMHO, foaf:membershipClass is kind of hacky. It requires non-trivial ontology-specific reasoning which can be replaced by standard owl semantics. Let's take the example on the webpage: ILRT staff Already it's using fairly complex owl-isms. Also, I do not think it means what the page thinks it means, but that's a seperate issues (all ILRT staff have a certain work homepage, but others may have that same work homepage and not be staff). To make it into owl-dl, we'll have to add one property - memberOf, and declare it to be the inverse of foaf:member. ILRT Staff Now we have that ILRTStaffGroup has all ILRTStaff as members, without having to define our own reasoning directives. -Ron Alford From julian_bond at voidstar.com Thu Sep 2 18:49:41 2004 From: julian_bond at voidstar.com (Julian Bond) Date: Mon Dec 18 17:23:56 2006 Subject: [rdfweb-dev] FOAF-Galway Message-ID: I'd just like to thank everyone and anyone involved in organising FOAF-Galway. I had a great time with a great bunch of people and learnt a lot. Well done. -- Julian Bond Email&MSM: julian.bond at voidstar.com Webmaster: http://www.ecademy.com/ Personal WebLog: http://www.voidstar.com/ M: +44 (0)77 5907 2173 T: +44 (0)192 0412 433 From nick.kings at bt.com Fri Sep 3 08:51:59 2004 From: nick.kings at bt.com (nick.kings@bt.com) Date: Mon Dec 18 17:23:56 2006 Subject: [rdfweb-dev] FOAF-Galway Message-ID: <21DA6754A9238B48B92F39637EF307FD0964EBB7@i2km41-ukdy.domain1.systemhost.net> I'd like to second that! Thanks!!! Nick _____ Nicholas J. Kings (Nick), Next Generation Web Research, BT Exact Technologies, http://www.btexact.com/ _____ > -----Original Message----- > From: rdfweb-dev-bounces@vapours.rdfweb.org > [mailto:rdfweb-dev-bounces@vapours.rdfweb.org] On Behalf Of > Julian Bond > Sent: 02 September 2004 19:50 > To: rdfweb-dev@vapours.rdfweb.org > Subject: [rdfweb-dev] FOAF-Galway > > I'd just like to thank everyone and anyone involved in > organising FOAF-Galway. I had a great time with a great bunch > of people and learnt a lot. Well done. > > -- > Julian Bond Email&MSM: julian.bond at voidstar.com > Webmaster: http://www.ecademy.com/ > Personal WebLog: http://www.voidstar.com/ > M: +44 (0)77 5907 2173 T: +44 (0)192 0412 433 > > _______________________________________________ > rdfweb-dev mailing list > rdfweb-dev@vapours.rdfweb.org > wiki: http://rdfweb.org/topic/FoafProject > http://rdfweb.org/mailman/listinfo/rdfweb-dev > From ldodds at ingenta.com Fri Sep 3 11:25:17 2004 From: ldodds at ingenta.com (Leigh Dodds) Date: Mon Dec 18 17:23:56 2006 Subject: [rdfweb-dev] [Fwd: FOAF IrcMeeting] Message-ID: <4138549D.80208@ingenta.com> resending as I was unsubscribed... -------- Original Message -------- Subject: FOAF IrcMeeting Date: Fri, 03 Sep 2004 10:40:57 +0100 From: Leigh Dodds To: rdfweb Comments re: regular FOAF IRC meetings. See also http://rdfweb.org/topic/IrcMeeting. To make the meetings most effective, can we agree on some basic procedures for carrying them out? - topics are published ~week in advance, ideally more so that comments from all communities can be included. - each chat is "lead" by a moderator, who is responsible for taking notes, and summarising the discussion. Perhaps the sponsor for a particular topic becomes the moderator. - each chat is summarised in a page in the Wiki, complete with list of actions/decisions - chat summary is circulated to this list and also linked from FOAF blog to ensure that people who couldn't attend (or don't use IRC) can be included. Common sense stuff I suppose, but doesn't hurt to write it down. L. From ldodds at ingenta.com Fri Sep 3 11:25:31 2004 From: ldodds at ingenta.com (Leigh Dodds) Date: Mon Dec 18 17:23:56 2006 Subject: [rdfweb-dev] [Fwd: Process, Issue Tracker] Message-ID: <413854AB.7070205@ingenta.com> resending... -------- Original Message -------- Subject: Process, Issue Tracker Date: Fri, 03 Sep 2004 10:50:22 +0100 From: Leigh Dodds To: rdfweb Reading through the FOAFCommunityProcess page on the Wiki [1] I agreed whole-heartedly with the point that notes that having a clear decision trail/record is required. I'm just wondering whether we can make some improvements to the IssueTracker page to make things easier to find (and fix). One improvement that I'd personally find useful is breaking down the "issues" into categories: spec revision proposals (e.g. new terms), spec maintenance issues (e.g. docs), new areas to explore (e.g. PGP vocab), etc. Some of these are things that we could collaborativel resolve, i.e. anyone could propose draft text to better document a term, or provide patches to the schema to ensure human and machine-readable versions are in accord. It's the other issues that are the ones that require an audit history: why did we add/remove a term, why was it changed, etc. And it's these issues that require the most discussion. Would help to be able to list them all separately. Maybe we just just move a bunch of items to a new page? (Although I did like the bugzilla interface as that allows searching for items by category) Cheers, L. [1]. http://rdfweb.org/topic/FOAFCommunityProcess From crschmidt at crschmidt.net Fri Sep 3 15:29:08 2004 From: crschmidt at crschmidt.net (Christopher Schmidt) Date: Mon Dec 18 17:23:56 2006 Subject: [rdfweb-dev] [Fwd: FOAF IrcMeeting] Message-ID: <20040903152908.GC5460@peanut.crschmidt.net> A few things from discussion today on IRC, as well as some responses. 1. Discussions on the topic list at IrcMeeting right now are foaf related, but really apply more generally to RDF. It might be better to hold the meetings in #rdfig, in terms of attracting more interested people as well as having a chump, keeping everyone involved, etc. 2. These meetings can (and probably will) be completely seperate from the FoafCommunityProcess meetings, designed for extension of the ontology: [[[ * agreed that a regular 'heartbeat' for status/progress updates to FOAF spec (issue lists, TODOs etc) would be helpful. * agreed an initial IRC meeting 1700pm BST 2004-09-15 in #foaf, for status/update from editor; danbri to circulate a meeting agenda to rdfweb-dev. ]]] -- http://rdfweb.org/topic/FOAFCommunityProcess Although these meetings fall into a similar arena, there is no need to discuss updates to the schema on a weekly basis. However, with a wide range of topics, weekly meetings along the lines of "interaction between $topic and FOAF" don't seem out of the question. > Comments re: regular FOAF IRC meetings. See also > http://rdfweb.org/topic/IrcMeeting. > > To make the meetings most effective, can we agree on > some basic procedures for carrying them out? > > - topics are published ~week in advance, ideally more > so that comments from all communities can be included. > > - each chat is "lead" by a moderator, who is responsible > for taking notes, and summarising the discussion. Perhaps > the sponsor for a particular topic becomes the moderator. > > - each chat is summarised in a page in the Wiki, complete with > list of actions/decisions > > - chat summary is circulated to this list and also linked from > FOAF blog to ensure that people who couldn't attend (or don't > use IRC) can be included. > > Common sense stuff I suppose, but doesn't hurt to write it down. > > L. I have no problems with any of these suggestions, and since I'm one to open my big mouth a lot and not do a lot about it, I'd be glad to act as a moderator for at least the first meeting. Since I do want to get this started, and not to interfere with the ontology discussion on the 15th, I'd like to propose that we wait until Noon GMT, Sunday to see if there's a definite preference for a topic. If there isn't, I'll make an executive decision as a moderator of the talk as to which topic will be discussed, in an effort to allow people to gather whatever thoughts they want to share. The agenda I had in mind would be something along the lines of the following (times in Eastern because that's what I'm thinking in): 11:00AM Call to Order, inform other people meeting is starting Introductions: Name, Rank, Serial Number 11:05AM Topic introduction, linking to relevant resources to get everyone involved on the same page 11:15AM Discussion: What people would like to see, what's needed, what's missing, what's okay but needs more work, general topic chatter 11:40AM Wrap up - move important/relevant links to wiki, add participants list. Define goals - what do people want to see, what needs to be done, "Action Items", also added to wiki 11:50AM Discussion of next meeting topic Then an email could be sent to the list detailing what people 1. Are working on 2. Will be working on 3. Discussed 4. Think about the topic Obviously, these times are relatively fluid, if the discussion needs or wants to continue people can either break off and work on it or something similar. I figure an hour is the longest a group of people can reasonably devote at random times of the day, although if it needs to go longer it can. Does this seem like a reasonable goal and timeline? Anyone have any objections? Anyone want to participate? :) The time of the meeting in many locations: http://tinyurl.com/65f37 -- Christopher Schmidt From Libby.Miller at bristol.ac.uk Sat Sep 4 19:32:48 2004 From: Libby.Miller at bristol.ac.uk (Libby Miller) Date: Mon Dec 18 17:23:56 2006 Subject: [rdfweb-dev] FOAF-Galway In-Reply-To: References: Message-ID: hey Julian! I'm glad you could make it and that you had a good time. It was great to see you :) Thanks to everyone who attended. I had a really productive and fun time, and I hope you did too (we'll give you the opportunity to comment on that soon I hope). Libby On Thu, 2 Sep 2004, Julian Bond wrote: > I'd just like to thank everyone and anyone involved in organising > FOAF-Galway. I had a great time with a great bunch of people and learnt a > lot. Well done. > > -- > Julian Bond Email&MSM: julian.bond at voidstar.com > Webmaster: http://www.ecademy.com/ > Personal WebLog: http://www.voidstar.com/ > M: +44 (0)77 5907 2173 T: +44 (0)192 0412 433 > > _______________________________________________ > rdfweb-dev mailing list > rdfweb-dev@vapours.rdfweb.org > wiki: http://rdfweb.org/topic/FoafProject > http://rdfweb.org/mailman/listinfo/rdfweb-dev > From Libby.Miller at bristol.ac.uk Sat Sep 4 19:40:16 2004 From: Libby.Miller at bristol.ac.uk (Libby Miller) Date: Mon Dec 18 17:23:56 2006 Subject: [rdfweb-dev] FOAF-Galway In-Reply-To: <21DA6754A9238B48B92F39637EF307FD0964EBB7@i2km41-ukdy.domain1.systemhost.net> References: <21DA6754A9238B48B92F39637EF307FD0964EBB7@i2km41-ukdy.domain1.systemhost.net> Message-ID: Thanks for coming Nick :) Libby On Fri, 3 Sep 2004 nick.kings@bt.com wrote: > > I'd like to second that! > > Thanks!!! > > Nick > > > > _____ > > Nicholas J. Kings (Nick), > Next Generation Web Research, BT Exact Technologies, > http://www.btexact.com/ > > _____ > > > >> -----Original Message----- >> From: rdfweb-dev-bounces@vapours.rdfweb.org >> [mailto:rdfweb-dev-bounces@vapours.rdfweb.org] On Behalf Of >> Julian Bond >> Sent: 02 September 2004 19:50 >> To: rdfweb-dev@vapours.rdfweb.org >> Subject: [rdfweb-dev] FOAF-Galway >> >> I'd just like to thank everyone and anyone involved in >> organising FOAF-Galway. I had a great time with a great bunch >> of people and learnt a lot. Well done. >> >> -- >> Julian Bond Email&MSM: julian.bond at voidstar.com >> Webmaster: http://www.ecademy.com/ >> Personal WebLog: http://www.voidstar.com/ >> M: +44 (0)77 5907 2173 T: +44 (0)192 0412 433 >> >> _______________________________________________ >> rdfweb-dev mailing list >> rdfweb-dev@vapours.rdfweb.org >> wiki: http://rdfweb.org/topic/FoafProject >> http://rdfweb.org/mailman/listinfo/rdfweb-dev >> > > _______________________________________________ > rdfweb-dev mailing list > rdfweb-dev@vapours.rdfweb.org > wiki: http://rdfweb.org/topic/FoafProject > http://rdfweb.org/mailman/listinfo/rdfweb-dev > From Libby.Miller at bristol.ac.uk Sun Sep 5 00:07:51 2004 From: Libby.Miller at bristol.ac.uk (Libby Miller) Date: Mon Dec 18 17:23:56 2006 Subject: [rdfweb-dev] more photos from foaf-galway Message-ID: http://swordfish.rdfweb.org/photos/2004/08/31/ http://swordfish.rdfweb.org/photos/2004/09/01/ http://swordfish.rdfweb.org/photos/2004/09/02/ http://swordfish.rdfweb.org/photos/2004/09/03/ Libby From danny.ayers at gmail.com Sun Sep 5 19:08:00 2004 From: danny.ayers at gmail.com (Danny Ayers) Date: Mon Dec 18 17:23:56 2006 Subject: [rdfweb-dev] Re: FoafInBrief (info towards Atom PersonConstruct) In-Reply-To: References: <1f2ed5cd04090509585ba3105b@mail.gmail.com> Message-ID: <1f2ed5cd0409051208b547e52@mail.gmail.com> On Sun, 05 Sep 2004 11:16:01 -0700, Tim Bray wrote: > On Sep 5, 2004, at 9:58 AM, Danny Ayers wrote: > In my experience FOAF is well thought-out and (speaking personally) I'd > be happy to avoid redesigning personal info in Atom by getting help > from FOAF. But author-identification is so basic to Atom that dipping > into another namespace seems egregious, so we'd want to use FOAF by > copy rather than by reference. Yep, that sounds reasonable. The path of least resistance being to learn what we can from FOAF's identification system, and implement something specific to Atom based on those lessons. My hope is that a side effect will be a fairly simple way of working with Atom and FOAF within the same application. Dan will hopefully fill in the details (he was flying to Tokyo today, so it probably won't be immediate), but an overhaul of the vocabulary and documentation was scheduled to follow the summer events - FOAFCamp in the Netherlands and a FOAF workshop in Galway. The latter only finished a couple of days go, but from what I gather a roadmap was on the agenda, and increased standardisation was under discussion. Cheers, Danny. -- http://dannyayers.com From danbri at w3.org Mon Sep 6 06:12:43 2004 From: danbri at w3.org (Dan Brickley) Date: Mon Dec 18 17:23:56 2006 Subject: [rdfweb-dev] Re: FoafInBrief (info towards Atom PersonConstruct) In-Reply-To: <1f2ed5cd0409051208b547e52@mail.gmail.com> References: <1f2ed5cd04090509585ba3105b@mail.gmail.com> <1f2ed5cd0409051208b547e52@mail.gmail.com> Message-ID: <20040906061243.GB7484@homer.w3.org> * Danny Ayers [2004-09-05 21:08+0200] > On Sun, 05 Sep 2004 11:16:01 -0700, Tim Bray wrote: > > On Sep 5, 2004, at 9:58 AM, Danny Ayers wrote: > > > In my experience FOAF is well thought-out and (speaking personally) I'd > > be happy to avoid redesigning personal info in Atom by getting help > > from FOAF. But author-identification is so basic to Atom that dipping > > into another namespace seems egregious, so we'd want to use FOAF by > > copy rather than by reference. > > Yep, that sounds reasonable. The path of least resistance being to > learn what we can from FOAF's identification system, and implement > something specific to Atom based on those lessons. My hope is that a > side effect will be a fairly simple way of working with Atom and FOAF > within the same application. > > Dan will hopefully fill in the details (he was flying to Tokyo today, > so it probably won't be immediate), but an overhaul of the vocabulary > and documentation was scheduled to follow the summer events - FOAFCamp > in the Netherlands and a FOAF workshop in Galway. The latter only > finished a couple of days go, but from what I gather a roadmap was on > the agenda, and increased standardisation was under discussion. Hi. Made it to Tokyo, though my brain isn't all here yet. I would be pleased to see a defined way of getting from Atom's notion of person to FOAF's, and back again. The core idea in FOAF for IDs is to use any properties that are uniquely identifying, rather than having silly angels-on-pinheads arguments about "what the URI is for a person". We say that 'homepage', 'weblog', (personal)'mbox' and 'mbox_sha1sum' are such things. There are also a few other pieces of FOAF that can play interestingly in an RSS1 environment, such as 'topic' (a relation between a document and something it is about); 'primaryTopic' (a relation between a topic and the main thing it is about), 'img' a relation between [from memory] a person and an image that is particularly characterstic of it. Also 'depicts' is popular for image description apps (moblogging etc), and relates an Image to a thing it depicts. These combine in ways that go a bit beyond people-description, eg. you can say that a Document has as its primaryTopic a thing that has a homepage of http://www.somemovie.example.com/. The idea is to emphasise idioms that allow for data merging across very loosly coordinated systems, by piggybacking off of very well known identifiers. Other person description things include workplaceHomepage, which is a basic construct that should allow cluetrainish apps such as 'find me all the things that are of type rss:channel and which are made by persons whose workplaceHomepage is http://www.sun.com/ or http://www.microsoft.com/. I doubt Atom'd want to go into such detail in its core, but as we get a process together (Galway notes are in http://rdfweb.org/topic/FOAFCommunityProcess), we will start tagging bits of FOAF as 'stable' and that might give groups such as Atom a bit more evidence that they're reasonable things to cite or copy. I think there's a good case for publishing something through W3C (eg. as an Interest Group Note of the SW IG) and/or a snapshot of term definitions in IETF somehow. I'm currently pretty lukewarm on the idea of transferring the whole thing to W3C, but have been discussing the idea of making a task force of the Semantic Web Best Practices WG that uses FOAF as a case study of how to (or not to!) make SW vocabs, and the issues that arise. Back re Atom, perhaps GRDDL, http://www.w3.org/TR/2004/NOTE-grddl-20040413/ would be a way of characterising a mapping from Atom's syntax into something RDF tools could eat... I do worry about people getting confused by seeing bits of FOAF RDF appear in non-RDF syntax, but if the FOAF subtrees of the Atom file used RDF notation, maybe that'd work. I'll have a think. Certainly for basic stuff like 'Person','mbox','mbox_sha1sum','homepage', 'weblog' there's a decent case for copy rather than reference. I need to have another look at latest Atom/person stuff before I can be any more useful here. It might not get gotten to until I'm back in Bristol on 14/15th... hope this helps, cheers, Dan ps. btw I'm not sure we say in the spec (we should) but FOAF owes a great deal to MCF, especially Guha and TimBray's MCF-in-XML spec... From henry.story at bblfish.net Mon Sep 6 15:23:26 2004 From: henry.story at bblfish.net (Henry Story) Date: Mon Dec 18 17:23:56 2006 Subject: [rdfweb-dev] Re: FoafInBrief (info towards Atom PersonConstruct) In-Reply-To: <20040906061243.GB7484@homer.w3.org> References: <1f2ed5cd04090509585ba3105b@mail.gmail.com> <1f2ed5cd0409051208b547e52@mail.gmail.com> <20040906061243.GB7484@homer.w3.org> Message-ID: First thanks a lot to Danny Ayers for the hard work he put into his clear and simple explanation of RDF and FOAF [1]. That is a very nice introduction to the technology. The best way to learn about it is to play with it a little. Making one's own FOAF file is a good first step. To those embarking down that road I seriously recommend writing the code in N3 and using a tool like cwm [2] to convert back and forth from that notation to the xml format. Take it from a newbie like me: it is not that difficult, and it's a lot of fun. On the topic at hand, I think there is a good beginning of a consensus being arrived at here: - the FOAF group is doing a lot of good work on the Person concept, - some core elements of that concept are required in Atom but importing all of the FOAF concepts might be overkill -> we just need to make sure we have a good interface between the Atom Person Construct and the FOAF equivalent (foaf:Agent or foaf:Person?) The above argument would tend to go against my model of Atom+OWL [3] where I directly import the FOAF ontology, and favor Danny's model [4] where he just defines the Person element inside of the atom namespace. In Java OO programming terms we can think of this as Danny creating a Person interface in the atom.* package, which all the other atom classes refer to when they want to speak of a Person object. What we need is a simple implementation class somewhere that implements that Person interface with the foaf equivalent. I would be really interested to know what the best way to do this in RDF is. One way to do this would be to have Atom:Person be a subclass of foaf:Person (or foaf:Agent?). Would there be better methods for this? I have set up a google newsgroup [5] for discussion of those issues that may appear to be more tangential to this group. Henry Story [1] http://www.intertwingly.net/wiki/pie/FoafInBrief [2] http://www.w3.org/2000/10/swap/doc/cwm.html [3] http://bblfish.net/work/atom-owl/2004-08-12/blogexample.html [4] http://semtext.org/atom/atom.html (section 3.2) [5] http://groups-beta.google.com/group/atom-owl From GK at ninebynine.org Mon Sep 6 15:22:15 2004 From: GK at ninebynine.org (Graham Klyne) Date: Mon Dec 18 17:23:56 2006 Subject: [rdfweb-dev] Postal address vocabulary In-Reply-To: <413323AF.6080503@gmuer.ch> Message-ID: <5.1.0.14.2.20040906153628.01fee8c8@127.0.0.1> At 14:55 30/08/04 +0200, Reto Bachmann-Gmuer wrote: >I've just added a proposal for a vocabulary to add postal addresses to >foaf-profiles at http://rdfweb.org/topic/AddressVocab > >Any comment would be appreciated! This appears to be a comprehensive approach, and I wonder if too comprehensive for a lightweight, general-use vocabulary like FOAF. ... For what it's worth, I informally proposed some FOAF terms for postal addresses some time ago, in support of some work I have been doing to use RDF in the generation of data for an IETF protocol element registry. An example of the terms I have used is: [[ hdr:Author_PR a wn:Person ; foaf:name "Peter W. Resnick" ; foaf:mbox ; foaf:organization "QUALCOMM Incorporated" ; foaf:workplacePostal [ foaf:street1 "5775 Morehouse Drive" ; foaf:city "San Diego" ; foaf:postcode "CA 92121-1714" ; foaf:country "USA" ] ; foaf:workplaceTel "+1 858 651 4478" ; foaf:workplaceFax "+1 858 651 1102" . ]] A more complete indication of the vocabulary might be gleaned from this query expression which is part of my software implementation: [[ @query hrep:HdrPersonPattern ( ?person ( foaf:name ?aname & ( foaf:mbox ?ambox | foaf:homepage ?ahomepage ) & [ foaf:organization ?orgname ] & [ @hrep:HdrPostalPattern ] & [ foaf:workplaceTel ?wtel ] & [ foaf:workplaceFax ?wfax ] & [ foaf:workplaceHomepage ?wurl ] ) ) @query hrep:HdrPostalPattern ( foaf:workplacePostal ?wp ( [ foaf:building ?wpbuilding ] & [ foaf:street1 ?wpstreet1 ] & [ foaf:street2 ?wpstreet2 ] & [ foaf:city ?wpcity ] & [ foaf:area ?wparea ] & [ foaf:postcode ?wppostcode ] & [ foaf:country ?wpcountry ] ) ) ]] I'm not suggesting this is any better than your proposal. I'm just mentioning it here because it was derived from a specific use-case (capturing details of IETF document editors), and as such might be a useful touchstone. #g ------------ Graham Klyne For email: http://www.ninebynine.org/#Contact From danny.ayers at gmail.com Mon Sep 6 17:07:31 2004 From: danny.ayers at gmail.com (Danny Ayers) Date: Mon Dec 18 17:23:56 2006 Subject: [rdfweb-dev] Re: FoafInBrief (info towards Atom PersonConstruct) In-Reply-To: References: <1f2ed5cd04090509585ba3105b@mail.gmail.com> <1f2ed5cd0409051208b547e52@mail.gmail.com> <20040906061243.GB7484@homer.w3.org> Message-ID: <1f2ed5cd040906100720259d6@mail.gmail.com> On Mon, 6 Sep 2004 17:23:26 +0200, Henry Story wrote: Thanks Henry, just a quick comment - > One way to do this would be to have Atom:Person be a subclass of > foaf:Person (or foaf:Agent?). Would there be better methods for this? I > have set up a google newsgroup [5] for discussion of those issues that > may appear to be more tangential to this group. Before anyone starts screaming, I would assume that the mapping will only be used in the RDF world, no dependencies would be introduced into Atom. Alternatives that spring to mind are stating that foaf:Person (/foaf:Agent) is an rdfs:subClassOf of atom:Person (the PersonConstruct), or each is an rdfs:subClassOf of the other (i.e. owl:equivalentClass). The current description of atom:Person is probably closer to foaf:Agent in that it may be an organisation. foaf:Agent is a little wider, it could also be something like a bot - maybe this broadening might be considered for the Atom construct. Cheers, Danny. -- http://dannyayers.com From marc at canter.com Tue Sep 7 14:27:33 2004 From: marc at canter.com (Marc Canter) Date: Mon Dec 18 17:23:56 2006 Subject: [rdfweb-dev] Re: FoafInBrief (info towards Atom PersonConstruct) In-Reply-To: Message-ID: <004401c494e6$d67b24c0$650a0a0a@Lucy> Connecting FOAF and Atom together is crucial. For the record - we've attempted to create a simplified early stage implementation of FOAF - called the FOAFnet.org. http://www.socialtext.net/foafnet/index.cgi In that spec - we start with a very simple iteration of FOAF, including ONLY: - name - face - email (sha1sum) - website - list of friends The goal of FOAFnet was to facilitate exchange of entire social networks. Ecademy demonstrated it working at FOAF Galway and Tribe.net will ship with it - in a couple of weeks. Whatever the Atom folks come up with - we can support in FOAF. Let's work towards smooth transparent interop - to make our end-user's lives better. - Marc Canter -----Original Message----- From: rdfweb-dev-bounces@vapours.rdfweb.org [mailto:rdfweb-dev-bounces@vapours.rdfweb.org] On Behalf Of Henry Story Sent: Monday, September 06, 2004 5:23 PM To: Atom Syntax Cc: foafnet@yahoogroups.com; rdfweb-dev@vapours.rdfweb.org; Tim Bray; Paul Hoffman / IMC Subject: [rdfweb-dev] Re: FoafInBrief (info towards Atom PersonConstruct) First thanks a lot to Danny Ayers for the hard work he put into his clear and simple explanation of RDF and FOAF [1]. That is a very nice introduction to the technology. The best way to learn about it is to play with it a little. Making one's own FOAF file is a good first step. To those embarking down that road I seriously recommend writing the code in N3 and using a tool like cwm [2] to convert back and forth from that notation to the xml format. Take it from a newbie like me: it is not that difficult, and it's a lot of fun. On the topic at hand, I think there is a good beginning of a consensus being arrived at here: - the FOAF group is doing a lot of good work on the Person concept, - some core elements of that concept are required in Atom but importing all of the FOAF concepts might be overkill -> we just need to make sure we have a good interface between the Atom Person Construct and the FOAF equivalent (foaf:Agent or foaf:Person?) The above argument would tend to go against my model of Atom+OWL [3] where I directly import the FOAF ontology, and favor Danny's model [4] where he just defines the Person element inside of the atom namespace. In Java OO programming terms we can think of this as Danny creating a Person interface in the atom.* package, which all the other atom classes refer to when they want to speak of a Person object. What we need is a simple implementation class somewhere that implements that Person interface with the foaf equivalent. I would be really interested to know what the best way to do this in RDF is. One way to do this would be to have Atom:Person be a subclass of foaf:Person (or foaf:Agent?). Would there be better methods for this? I have set up a google newsgroup [5] for discussion of those issues that may appear to be more tangential to this group. Henry Story [1] http://www.intertwingly.net/wiki/pie/FoafInBrief [2] http://www.w3.org/2000/10/swap/doc/cwm.html [3] http://bblfish.net/work/atom-owl/2004-08-12/blogexample.html [4] http://semtext.org/atom/atom.html (section 3.2) [5] http://groups-beta.google.com/group/atom-owl _______________________________________________ rdfweb-dev mailing list rdfweb-dev@vapours.rdfweb.org wiki: http://rdfweb.org/topic/FoafProject http://rdfweb.org/mailman/listinfo/rdfweb-dev From mof-rdf at mfd-consult.dk Tue Sep 7 19:51:53 2004 From: mof-rdf at mfd-consult.dk (Morten Frederiksen) Date: Mon Dec 18 17:23:56 2006 Subject: [rdfweb-dev] more photos from foaf-galway In-Reply-To: References: Message-ID: <200409072151.53369.mof-rdf@mfd-consult.dk> Hi all, Just a quick note of thanks to all at FOAF Galway, but of course especially to the organisers, and a pointer to a few more photos -- annotations are in the queue... http://www.wasab.dk/morten/2004/08/photos/eire/ Regards, Morten From gk at ninebynine.org Wed Sep 8 15:42:08 2004 From: gk at ninebynine.org (Graham Klyne) Date: Mon Dec 18 17:23:56 2006 Subject: [rdfweb-dev] Haskell (FOAFcamp follow-up) Message-ID: <5.1.0.14.2.20040908160809.02a41ea0@127.0.0.1> Following my brief show'n'tell at the recent FOAFcamp of a "literate Haskell" work-in-progress for exploring description logic reasoning [1], Tony Bowden has passed me a reference to a description of an experiment in which the US military surveyed a range of programming lanugages, and didn't believe that the Haskell solution was real: It is significant that Mr. Domanski, Mr. Banowetz and Dr. Brosgol were all surprised and suspicious when we told them that Haskell prototype P1 (see appendix B) is a complete tested executable program. We provided them with a copy of P1 without explaining that it was a program, and based on preconceptions from their past experience, they had studied P1 under the assumption that it was a mixture of requirements specification and top level design. They were convinced it was incomplete because it did not address issues such as data structure design and execution order." "Haskell vs. Ada vs. C++ vs. Awk vs. ..., An Experiment in Software Prototyping Productivity" http://www.haskell.org/papers/NSWC/jfp.ps If/when XML and RDF library support for Haskell is stable and easily available, I'd hope for some similar advantages from Haskell for semantic web application development. #g -- [1] http://www.ninebynine.org/Software/HaskellDL/DLExploration.lhs ------------ Graham Klyne For email: http://www.ninebynine.org/#Contact From danny.ayers at gmail.com Mon Sep 13 11:26:31 2004 From: danny.ayers at gmail.com (Danny Ayers) Date: Mon Dec 18 17:23:56 2006 Subject: [rdfweb-dev] RDF model sync, HTTP deltas, syndicated feeds (and a bounty) Message-ID: <1f2ed5cd0409130426540c86f9@mail.gmail.com> Long story short, PubSub.com is willing to offer a $5,000 bounty to whoever builds the "best" implementation of RFC3229 (Delta encoding in HTTP) with appropriate extensions for use with syndication feeds as an open source Apache module by Jan 1, 2005. The requirements for feed-oriented delta transfers are very, very close to those for passing around potentially huge, dynamically changing RDF/XML files, effectively synchronizing models between systems. Seems to me PubSub's solution could feed two birds with one bean, hence this mail. In the past few days there have been a couple of related developments - MSDN got stung with excessive bandwidth costs trying to broadcast the aggregated content of 1300+ feeds [2], and a (relatively!) simple algorithm for a delta-based solution has been described [3]. Most of the discussion has been taking place on the atom-syntax list [4]. Bob Wyman (of PubSub) describes the feed-oriented requirements some more at [5], there's a little more background on my blog at [6],[7]. Cheers, Danny. [1] http://www.ietf.org/rfc/rfc3229.txt [2] http://radio.weblogs.com/0001011/2004/09/08.html#a8195 [3] http://www.intertwingly.net/blog/2004/09/11/Vary-ETag [4] http://www.imc.org/atom-syntax/ [5] http://bobwyman.pubsub.com/main/2004/09/using_rfc3229_w.html [6] http://dannyayers.com/archives/2004/09/12/passing-things-around/ [7] http://dannyayers.com/archives/2004/09/13/geronimo/ -- http://dannyayers.com From reto at gmuer.ch Mon Sep 13 15:32:39 2004 From: reto at gmuer.ch (Reto Bachmann-Gmuer) Date: Mon Dec 18 17:23:56 2006 Subject: [rdfweb-dev] Postal address vocabulary In-Reply-To: <5.1.0.14.2.20040906153628.01fee8c8@127.0.0.1> References: <5.1.0.14.2.20040906153628.01fee8c8@127.0.0.1> Message-ID: <4145BD97.7030501@gmuer.ch> Hi Graham, >> foaf-profiles at http://rdfweb.org/topic/AddressVocab > This appears to be a comprehensive approach, and I wonder if too > comprehensive for a lightweight, general-use vocabulary like FOAF. It's true that it long, I think a subset of the properties could be defined wich would be enough for most addresses. Maybe complexity is not foaf-style but I like the following aspects which I think are foafish: - decentral: separating post-operator information from the "objective" location makes it possible to be used independently of controlled systems. - culturale independence (this comes from copying from the S42 standard): any address can be represented and mechanism for locale specific rendering exist. - suitable for inference: it is clear if two persons live in the same house or street which may be hard if this information is joined in one literal. Of course aspects like sorting by street may be quite irrelevant in the foaf-world, but I think if complexity remains somehow reasonable it makes good sense to have one vocabulary for addresses which can be used in different areas (foaf, data-mining, post, geo) rather than multiple address-schemas. > > ... > > For what it's worth, I informally proposed some FOAF terms for postal > addresses some time ago, in support of some work I have been doing to > use RDF in the generation of data for an IETF protocol element > registry. An example of the terms I have used is: > [[ > hdr:Author_PR a wn:Person ; > foaf:name "Peter W. Resnick" ; > foaf:mbox ; > foaf:organization "QUALCOMM Incorporated" ; > foaf:workplacePostal > [ foaf:street1 "5775 Morehouse Drive" ; > foaf:city "San Diego" ; > foaf:postcode "CA 92121-1714" ; > foaf:country "USA" ] ; > foaf:workplaceTel "+1 858 651 4478" ; > foaf:workplaceFax "+1 858 651 1102" . > ]] I see that you include both organization and individual, which makes sense. I think that foaf should offer a vocabulary to express this, the address vocabulary I propopose for this at http://rdfweb.org/topic/RoleVocab. The idea is that an Agent participates in a Group having a certain Role in this group and that this Participation is itself a foaf:Agent. Like this the Participation could could be the adresse of an Adress, and of course have things like foaf:phone, foaf:knows which may or may not be the same as those of the individual or those of the group. I think it would be a pity if foaf would model an old-fashioned one person - one workplace world. cheers, reto From gk at ninebynine.org Mon Sep 13 19:08:56 2004 From: gk at ninebynine.org (Graham Klyne) Date: Mon Dec 18 17:23:56 2006 Subject: [rdfweb-dev] Postal address vocabulary In-Reply-To: <4145BD97.7030501@gmuer.ch> References: <5.1.0.14.2.20040906153628.01fee8c8@127.0.0.1> <5.1.0.14.2.20040906153628.01fee8c8@127.0.0.1> Message-ID: <5.1.0.14.2.20040913200604.02b9aee0@127.0.0.1> Reto, I acknowledge all the points you make in your response. My question (and this is a question, not an assertion rhetorically styled as a question) is this: should this kind of detailed address-description vocabulary be part of FOAF, or a separate vocabulary (e.g. another namespace, which could, of course, be used *with* FOAF)? #g -- At 17:32 13/09/04 +0200, Reto Bachmann-Gmuer wrote: >Hi Graham, > >>>foaf-profiles at http://rdfweb.org/topic/AddressVocab > >>This appears to be a comprehensive approach, and I wonder if too >>comprehensive for a lightweight, general-use vocabulary like FOAF. >It's true that it long, I think a subset of the properties could be >defined wich would be enough for most addresses. > >Maybe complexity is not foaf-style but I like the following aspects >which I think are foafish: >- decentral: separating post-operator information from the "objective" >location makes it possible to be used independently of controlled systems. >- culturale independence (this comes from copying from the S42 >standard): any address can be represented and mechanism for locale >specific rendering exist. >- suitable for inference: it is clear if two persons live in the same >house or street which may be hard if this information is joined in one >literal. > >Of course aspects like sorting by street may be quite irrelevant in the >foaf-world, but I think if complexity remains somehow reasonable it >makes good sense to have one vocabulary for addresses which can be used >in different areas (foaf, data-mining, post, geo) rather than multiple >address-schemas. > >>... >>For what it's worth, I informally proposed some FOAF terms for postal >>addresses some time ago, in support of some work I have been doing to use >>RDF in the generation of data for an IETF protocol element registry. An >>example of the terms I have used is: >>[[ >>hdr:Author_PR a wn:Person ; >> foaf:name "Peter W. Resnick" ; >> foaf:mbox ; >> foaf:organization "QUALCOMM Incorporated" ; >> foaf:workplacePostal >> [ foaf:street1 "5775 Morehouse Drive" ; >> foaf:city "San Diego" ; >> foaf:postcode "CA 92121-1714" ; >> foaf:country "USA" ] ; >> foaf:workplaceTel "+1 858 651 4478" ; >> foaf:workplaceFax "+1 858 651 1102" . >>]] >I see that you include both organization and individual, which makes >sense. I think that foaf should offer a vocabulary to express this, the >address vocabulary I propopose for this at >http://rdfweb.org/topic/RoleVocab. The idea is that an Agent participates >in a Group having a certain Role in this group and that this Participation >is itself a foaf:Agent. Like this the Participation could could be the >adresse of an Adress, and of course have things like foaf:phone, >foaf:knows which may or may not be the same as those of the individual or >those of the group. I think it would be a pity if foaf would model an >old-fashioned one person - one workplace world. > >cheers, >reto ------------ Graham Klyne For email: http://www.ninebynine.org/#Contact From stpeter at jabber.org Mon Sep 13 21:50:11 2004 From: stpeter at jabber.org (Peter Saint-Andre) Date: Mon Dec 18 17:23:56 2006 Subject: [rdfweb-dev] Re: RDF model sync, HTTP deltas, syndicated feeds (and a bounty) References: <1f2ed5cd0409130426540c86f9@mail.gmail.com> Message-ID: In article <1f2ed5cd0409130426540c86f9@mail.gmail.com>, Danny Ayers wrote: > http://bobwyman.pubsub.com/main/2004/09/using_rfc3229_w.html Bob also mentions that a true pubsub approach would be best in the long term; for instance, see: http://www.ietf.org/internet-drafts/draft-saintandre-atompub-notify-01.tx t /psa From zednenem at psualum.com Tue Sep 14 02:38:11 2004 From: zednenem at psualum.com (David Menendez) Date: Mon Dec 18 17:23:56 2006 Subject: [rdfweb-dev] Postal address vocabulary In-Reply-To: <5.1.0.14.2.20040913200604.02b9aee0@127.0.0.1> Message-ID: Graham Klyne writes: > I acknowledge all the points you make in your response. My question > (and this is a question, not an assertion rhetorically styled as a > question) is this: should this kind of detailed address-description > vocabulary be part of FOAF, or a separate vocabulary (e.g. another > namespace, which could, of course, be used *with* FOAF)? I'd make it a separate vocabulary (call it an adjunct or a satellite vocabulary). FOAF and the address vocab are both large enough that they'd be easier to manage separately, and making the address vocab separate might make it more attractive to other projects. The only negative point which comes to mind is that it's slightly more work for people generating FOAF profiles by hand. -- David Menendez | "In this house, we obey the laws | of thermodynamics!" From julian_bond at voidstar.com Tue Sep 14 07:27:21 2004 From: julian_bond at voidstar.com (Julian Bond) Date: Mon Dec 18 17:23:56 2006 Subject: [rdfweb-dev] Postal address vocabulary In-Reply-To: <5.1.0.14.2.20040913200604.02b9aee0@127.0.0.1> References: <5.1.0.14.2.20040906153628.01fee8c8@127.0.0.1> <5.1.0.14.2.20040906153628.01fee8c8@127.0.0.1> <4145BD97.7030501@gmuer.ch> <5.1.0.14.2.20040913200604.02b9aee0@127.0.0.1> Message-ID: Graham Klyne wrote: >I acknowledge all the points you make in your response. My question >(and this is a question, not an assertion rhetorically styled as a >question) is this: >should this kind of detailed address-description vocabulary be part of >FOAF, or a separate vocabulary (e.g. another namespace, which could, of >course, be used *with* FOAF)? I consider a generalised RDF (and/or XML) Address schema to be so important that it's development *must* be done in the context of a major standards organization. Given the history, rfc status and wide deployment and implementation of vCard I find it pretty hard to argue for a new one. So maybe what's really needed is a new vCard RFC that supersedes RFC 2426. If the problem is with the RDF representation of vCard, then that's being done within the W3C. If it's wrong there are well established ways to improve it. So I'm not against new projects to replace and redefine the same areas as existing schemas. But for something with such wide use I'd want to see a lot of justification and at least some comment about how you expect to get traction for the new standard. There's already a lot of FOAF out there that contains data in the vcard schema as well as many other schemas. And this is a key benefit of RDF. So it really doesn't look like there's a need to keep adding to the FOAF schema tags that overlap with other existing schemas. I've argued before that actually FOAF should be slimmed down so that it becomes a structural schema with only just enough data tags to make the structure useful and readable. Taken to an extreme (ie too far!) that would mean foaf:Person, foaf:mbox_sha1sum, foaf:name, foaf:knows and that's about all. Now that's clearly a mistake, but I am coming round to the idea that we shouldn't confuse the FOAF schema as structure by adding more and more general purpose tags that happen to be missing from the existing schema space. Which is all a long winded way of saying, if you see a hole create a new namespace. And then fight for mindshare for that namespace in the ecology of namespace standards. David Menendez wrote: >The only negative point which comes to mind is that it's slightly more >work for people generating FOAF profiles by hand. The only people generating FOAF by hand should be experimenters and programmer/implementors with an intimate knowledge of RDF *and nobody else*. If you're not creating tools, use the tools, Luke. -- Julian Bond Email&MSM: julian.bond at voidstar.com Webmaster: http://www.ecademy.com/ Personal WebLog: http://www.voidstar.com/ M: +44 (0)77 5907 2173 T: +44 (0)192 0412 433 From reto at gmuer.ch Tue Sep 14 12:57:57 2004 From: reto at gmuer.ch (Reto Bachmann-Gmuer) Date: Mon Dec 18 17:23:56 2006 Subject: [rdfweb-dev] Role-Vocab and Re: Postal address vocabulary In-Reply-To: References: Message-ID: <4146EAD5.3000908@gmuer.ch> David Menendez wrote: > Graham Klyne writes: > > >>I acknowledge all the points you make in your response. My question >>(and this is a question, not an assertion rhetorically styled as a >>question) is this: should this kind of detailed address-description >>vocabulary be part of FOAF, or a separate vocabulary (e.g. another >>namespace, which could, of course, be used *with* FOAF)? > > > I'd make it a separate vocabulary (call it an adjunct or a satellite > vocabulary). FOAF and the address vocab are both large enough that > they'd be easier to manage separately, and making the address vocab > separate might make it more attractive to other projects. > > The only negative point which comes to mind is that it's slightly more > work for people generating FOAF profiles by hand. I'd suggest to have the address vocab a sperate vocabulary except maybe the #address property which has domain foaf:Agent. Have you had a look at the role vocab at http://rdfweb.org/topic/RoleVocab? I think this could usefully be part of the foaf-core vocab: seing foaf primarily as a vocabulary for describing (social) agents and their relations. Possibly this would make foaf:workplaceHompage obsolete. I've put an example on the wiki page showing a more comprehensive version of the workplaceHomepage example from the foaf-spec. cheers, reto From zednenem at psualum.com Wed Sep 15 01:13:16 2004 From: zednenem at psualum.com (David Menendez) Date: Mon Dec 18 17:23:56 2006 Subject: [rdfweb-dev] Postal address vocabulary In-Reply-To: Message-ID: Julian Bond writes: > Graham Klyne wrote: > >I acknowledge all the points you make in your response. My question > >(and this is a question, not an assertion rhetorically styled as a > >question) is this: should this kind of detailed address-description > >vocabulary be part of FOAF, or a separate vocabulary (e.g. another > >namespace, which could, of course, be used *with* FOAF)? > > I consider a generalised RDF (and/or XML) Address schema to be so > important that it's development *must* be done in the context of a > major standards organization. Given the history, rfc status and wide > deployment and implementation of vCard I find it pretty hard to argue > for a new one. So maybe what's really needed is a new vCard RFC that > supersedes RFC 2426. If the problem is with the RDF representation of > vCard, then that's being done within the W3C. If it's wrong there are > well established ways to improve it. Without agreeing or disagreeing, I'll note: (1) The only RDF schema for vCard I can find is a W3C Technical Note[1] from 2001 (which, in my opinion, could use some refactoring) (2) Reto's proposal[2] is an RDF translation of S42-3 from the Univeral Postal Union[3]. (3) S42-3 deals specifically with postal addresses (it represents the information on an individual envelopes), while vCard has a wider focus (it represents the information one might have on a business card). > David Menendez wrote: > >The only negative point which comes to mind is that it's slightly > >more work for people generating FOAF profiles by hand. > > The only people generating FOAF by hand should be experimenters and > programmer/implementors with an intimate knowledge of RDF *and nobody > else*. If you're not creating tools, use the tools, Luke. Exactly. My point was that the only argument I could think of for adding a full-blown address vocabulary to FOAF was very weak. [1] [2] [3] -- David Menendez From mcculley at stackframe.com Wed Sep 15 04:15:58 2004 From: mcculley at stackframe.com (Gene McCulley) Date: Mon Dec 18 17:23:56 2006 Subject: [rdfweb-dev] www.foafspace.com Message-ID: RDF folks, I have built a foaf search engine and put it at www.foafspace.com. I have not yet finished the smushing mechanism, so the engine currently returns more entries than desired. It still has some problems with the occasional RDF file. Other than that, it is mostly stable. I would appreciate any feedback regarding desired features and bugs found. Thanks. -------------- next part -------------- A non-text attachment was scrubbed... Name: PGP.sig Type: application/pgp-signature Size: 186 bytes Desc: This is a digitally signed message part Url : http://lists.usefulinc.com/pipermail/foaf-dev/attachments/20040915/306dcc2b/PGP-0001.pgp From julian_bond at voidstar.com Wed Sep 15 06:51:22 2004 From: julian_bond at voidstar.com (Julian Bond) Date: Mon Dec 18 17:23:56 2006 Subject: [rdfweb-dev] www.foafspace.com In-Reply-To: References: Message-ID: Gene McCulley wrote: >I have built a foaf search engine and put it at www.foafspace.com. > >I have not yet finished the smushing mechanism, so the engine currently >returns more entries than desired. It still has some problems with the >occasional RDF file. Other than that, it is mostly stable. > >I would appreciate any feedback regarding desired features and bugs >found. Nice. What's the underlying technology? And what Schemas does it understand? -- Julian Bond Email&MSM: julian.bond at voidstar.com Webmaster: http://www.ecademy.com/ Personal WebLog: http://www.voidstar.com/ M: +44 (0)77 5907 2173 T: +44 (0)192 0412 433 From john.breslin at deri.org Wed Sep 15 09:46:17 2004 From: john.breslin at deri.org (John Breslin) Date: Mon Dec 18 17:23:56 2006 Subject: [rdfweb-dev] www.foafspace.com In-Reply-To: Message-ID: <41481D79.1154.4DF07A@localhost> Looks cool, just submitted my FOAF profile from boards.jp and it spidered 364 other members from that. Are you just using the unofficial foaf:dateOfBirth or any other variations? J. -- Dr. John Breslin Digital Enterprise Research Institute http://www.johnbreslin.com/ john.breslin@deri.org From mcculley at stackframe.com Wed Sep 15 12:56:17 2004 From: mcculley at stackframe.com (Gene McCulley) Date: Mon Dec 18 17:23:56 2006 Subject: [rdfweb-dev] www.foafspace.com In-Reply-To: References: Message-ID: It uses a Java web application engine that I built for other purposes. The spidering and database management are done in Java as well. I wrote my own library for doing FOAF/RDF parsing, which might not be the right thing to do. I intend to take a look at Jena to see if I should use that for the RDF bits. It currently understands most of the foaf schema. I intend to add geo and vcard support. On Sep 15, 2004, at 2:51 AM, Julian Bond wrote: > Nice. What's the underlying technology? And what Schemas does it > understand? -------------- next part -------------- A non-text attachment was scrubbed... Name: PGP.sig Type: application/pgp-signature Size: 186 bytes Desc: This is a digitally signed message part Url : http://lists.usefulinc.com/pipermail/foaf-dev/attachments/20040915/8b667645/PGP-0001.pgp From mcculley at stackframe.com Wed Sep 15 13:05:11 2004 From: mcculley at stackframe.com (Gene McCulley) Date: Mon Dec 18 17:23:56 2006 Subject: [rdfweb-dev] www.foafspace.com In-Reply-To: <41481D79.1154.4DF07A@localhost> References: <41481D79.1154.4DF07A@localhost> Message-ID: Thanks for submitting the boards.jp link. If anyone has pointers to other sites that have multiple FOAF entries (that aren't shown on http://www.foafspace.com/hosts), I would appreciate it. I am using the unofficial dateOfBirth just because that is what LiveJournal emits. I tried to stay away from the bits of the schema not marked stable, but there is a lot of content out there using unofficial tags. I see that some files use the vCard BDAY or a birthday field in the bio schema. I intend to add support for those. On Sep 15, 2004, at 5:46 AM, John Breslin wrote: > Looks cool, just submitted my FOAF profile from boards.jp and it > spidered 364 > other members from that. > > Are you just using the unofficial foaf:dateOfBirth or any other > variations? > > J. > -- > Dr. John Breslin > Digital Enterprise Research Institute > http://www.johnbreslin.com/ > john.breslin@deri.org > > -------------- next part -------------- A non-text attachment was scrubbed... Name: PGP.sig Type: application/pgp-signature Size: 186 bytes Desc: This is a digitally signed message part Url : http://lists.usefulinc.com/pipermail/foaf-dev/attachments/20040915/47a5153b/PGP-0001.pgp From richard at cyganiak.de Wed Sep 15 21:57:50 2004 From: richard at cyganiak.de (Richard Cyganiak) Date: Mon Dec 18 17:23:56 2006 Subject: [rdfweb-dev] Issue: FOAF RDFS should not import RDFS and OWL vocabularies Message-ID: <49882A1F-0762-11D9-9A5B-000A95DABFDC@cyganiak.de> Hi everyone, the FOAF RDFS file[1] imports the RDFS and OWL vocabulary files: The OWL guide[2] recommends against doing that: Note that in order to use the OWL vocabulary you do not need to import the owl.rdf ontology. In fact, such an import is not recommended. I've added this to the IssueTracker[3]. It's not a big issue, but I thought I'd mention it anyway. Lots of ontologies/vocabularies do that. Is there a good reason that I am missing? Richard [1] http://xmlns.com/foaf/0.1/index.rdf [2] http://www.w3.org/TR/owl-guide/ [3] http://rdfweb.org/topic/OwlImportsIssue From danbri at w3.org Wed Sep 15 23:46:11 2004 From: danbri at w3.org (Dan Brickley) Date: Mon Dec 18 17:23:56 2006 Subject: [rdfweb-dev] Issue: FOAF RDFS should not import RDFS and OWL vocabularies In-Reply-To: <49882A1F-0762-11D9-9A5B-000A95DABFDC@cyganiak.de> References: <49882A1F-0762-11D9-9A5B-000A95DABFDC@cyganiak.de> Message-ID: <20040915234610.GA18782@homer.w3.org> * Richard Cyganiak [2004-09-15 23:57+0200] > Hi everyone, > > the FOAF RDFS file[1] imports the RDFS and OWL vocabulary files: > > > > > The OWL guide[2] recommends against doing that: > > Note that in order to use the OWL vocabulary you do not need to > import the owl.rdf ontology. In fact, such an import is not > recommended. > > I've added this to the IssueTracker[3]. It's not a big issue, but I > thought I'd mention it anyway. Lots of ontologies/vocabularies do that. > Is there a good reason that I am missing? I added the owl:imports statements (with some skepticism) pretty early on in OWL's lifetime, I think (FOAF predates OWL). I'd be quite happy to remove these statements, and perhaps just add rdf:seeAlso crossrefs to provide a hint for crawlers (since multilingual labels can be found via crawling from the rdfs namespace). owl:imports is imho a bit wierd, sort of mixes up the layers in a strange way, and doesn't add much anyway... If anyone objects to removing these triples, now would be a great time to say! Dan > > Richard > > [1] http://xmlns.com/foaf/0.1/index.rdf > [2] http://www.w3.org/TR/owl-guide/ > [3] http://rdfweb.org/topic/OwlImportsIssue > > > _______________________________________________ > rdfweb-dev mailing list > rdfweb-dev@vapours.rdfweb.org > wiki: http://rdfweb.org/topic/FoafProject > http://rdfweb.org/mailman/listinfo/rdfweb-dev From ronwalf at volus.net Thu Sep 16 01:30:14 2004 From: ronwalf at volus.net (Ron Alford) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] Issue: FOAF RDFS should not import RDFS and OWL vocabularies In-Reply-To: <20040915234610.GA18782@homer.w3.org> References: <49882A1F-0762-11D9-9A5B-000A95DABFDC@cyganiak.de> <20040915234610.GA18782@homer.w3.org> Message-ID: <4148ECA6.3030002@volus.net> Dan Brickley wrote: > I added the owl:imports statements (with some skepticism) pretty early > on in OWL's lifetime, I think (FOAF predates OWL). I'd be quite happy to > remove these statements, and perhaps just add rdf:seeAlso crossrefs to > provide a hint for crawlers (since multilingual labels can be found via > crawling from the rdfs namespace). Yeah, a seeAlso would be fine there, and probably more along the lines of what you mean. > owl:imports is imho a bit wierd, sort of mixes up the layers in a > strange way, and doesn't add much anyway... owl:imports is only really needed at the ontology level if you are really making use of the structural parts of another ontology. What it's really useful for is in instance data. I use it over at http://www.mindswap.org/2004/owl/mindswappers to point to an organization-specific extension of foaf (which happens to point to a dl-friendly version of the foaf ont). The owl:imports has well defined semantics with respect to reasoning. This makes it easy for a spidering owl agent to perform inferences over a file without having to already know about all the schemas the data could be using. > > If anyone objects to removing these triples, now would be a great time > to say! > Removing them would be great! I'm glad to see some interest from someone other than myself about making foaf more owl-friendly! -Ron From Libby.Miller at bristol.ac.uk Thu Sep 16 09:25:18 2004 From: Libby.Miller at bristol.ac.uk (Libby Miller) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] Issue: FOAF RDFS should not import RDFS and OWL vocabularies In-Reply-To: <49882A1F-0762-11D9-9A5B-000A95DABFDC@cyganiak.de> References: <49882A1F-0762-11D9-9A5B-000A95DABFDC@cyganiak.de> Message-ID: Thanks for this. I think someone else mentionned it (possibly in conversation) so it's good to have it documented. I'm not sure why the issue is important, though... Libby On Wed, 15 Sep 2004, Richard Cyganiak wrote: > Hi everyone, > > the FOAF RDFS file[1] imports the RDFS and OWL vocabulary files: > > > > > The OWL guide[2] recommends against doing that: > > Note that in order to use the OWL vocabulary you do not need to > import the owl.rdf ontology. In fact, such an import is not > recommended. > > I've added this to the IssueTracker[3]. It's not a big issue, but I thought > I'd mention it anyway. Lots of ontologies/vocabularies do that. Is there a > good reason that I am missing? > > Richard > > [1] http://xmlns.com/foaf/0.1/index.rdf > [2] http://www.w3.org/TR/owl-guide/ > [3] http://rdfweb.org/topic/OwlImportsIssue > > > _______________________________________________ > rdfweb-dev mailing list > rdfweb-dev@vapours.rdfweb.org > wiki: http://rdfweb.org/topic/FoafProject > http://rdfweb.org/mailman/listinfo/rdfweb-dev > From jo at abduction.org Thu Sep 16 15:11:12 2004 From: jo at abduction.org (Jo Walsh) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] Re: [Geowanking] next gen geoURL In-Reply-To: References: <6.0.0.22.2.20040916120413.01b3b168@sunrise.sli.unimelb.edu.au> <753344F9-078F-11D9-96C4-000A95DA2B18@eogeo.org> <453B13F6-0791-11D9-8346-000A95C54164@splattercast.com> Message-ID: <20040916151112.GC23473@vishnu.tridity.org> hey allan, geowankers, On Wed, Sep 15, 2004 at 11:52:49PM -0400, Allan Doyle wrote: > For all the people who are chafing because what's out there is not > quite sufficient in some dimension. One of my particular goals is the > ability to put metadata about geographic data into RSS so small > organizations can advertise their data, or can advertise their need for > data. Aggregators can pick those things up and start making matches > among the haves and the have nots. i wrote a quick braindump, couple of months back, for the craigslist techs about geoannotating rss, html etc with rdf. might be of relevance. http://frot.org/geo/craigslist.html having a URI scheme, or set of them, to approximately indicate locales or places, can be as useful as coordinates and i think will come up a lot during the w3 budapest workshop early next month. this approach to rdf-in-html looks nasty because of the necessity of keeping the XML namespace headers. it's not really human-writable; this discussion seems to be calling for human-writability. in an RSS1 context, not such a problem. i think the w3 is discussing this somewhere, but i'm not sure where. i think mike posted here, at the time, my naive attempts to turn the RDFIG geo namespace into an IETF draft, i never pushed this forward anywhere, but i think it encapsulates the 'decimal degrees in WGS84 for simplicity's sake' argument. http://space.frot.org/draft-geo-draft.html - feedback welcomed. -jo From wolf at bluehands.de Thu Sep 16 16:20:05 2004 From: wolf at bluehands.de (Heiner Wolf) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] Re: [Geowanking] next gen geoURL Message-ID: <11D121AB355B69448D3A9F2132D2A3AA42016B@niobe.BlueHands.de> Hi, do you know the IETF effort there. It is on the IETF standards track. Recent action: "Protocol Action: 'A Presence-based GEOPRIV Location Object Format' to Proposed Standard" Maybe it's useful for FOAF-RDF: See: Title : A Presence-based GEOPRIV Location Object Format Author(s) : J. Peterson Filename : draft-ietf-geopriv-pidf-lo-03.txt Pages : 24 Date : 2004-9-10 This draft is a work item of the Geographic Location/Privacy Working Group of the IETF. This document describes an object format for carrying geographical information on the Internet. This location object extends the Presence Information Data Format (PIDF), which was designed for communicating privacy-sensitive presence information and which has similar properties. A URL for this Internet-Draft is: http://www.ietf.org/internet-drafts/draft-ietf-geopriv-pidf-lo-03.txt hw -- Dr. Klaus H. Wolf bluehands GmbH & Co.mmunication KG http://www.bluehands.de/people/hw +49 (0721) 16108 75 -- Jabber enabled Virtual Presence on the Web: http://www.lluna.de/ Open Source Future History: http://www.galactic-developments.com/ From richard at cyganiak.de Thu Sep 16 17:07:33 2004 From: richard at cyganiak.de (Richard Cyganiak) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] Issue (?): URI of owl:Ontology header in the RDFS spec Message-ID: Hi everyone, I noticed another thing about the FOAF RDFS file. I'm not sure if I have a point this time or if it's just nitpicking. The URL of the RDFS file is http://xmlns.com/foaf/0.1/index.rdf . The owl:Ontology header of this file says: References: <49882A1F-0762-11D9-9A5B-000A95DABFDC@cyganiak.de> <49882A1F-0762-11D9-9A5B-000A95DABFDC@cyganiak.de> Message-ID: <5.1.0.14.2.20040916175728.02aa97b8@127.0.0.1> At 10:25 16/09/04 +0100, Libby Miller wrote: >Thanks for this. I think someone else mentionned it (possibly in >conversation) so it's good to have it documented. I'm not sure why the >issue is important, though... Could it be something to do with this... [[ The meaning of an occurrence of a URI is thus determined by its context, which we take to mean the document in which it appears, plus other documents explicitly mentioned in constructs like the OWL importing mechanism. ]] -- http://www-db.research.bell-labs.com/user/pfps/publications/meaning.pdf ? #g ------------ Graham Klyne For email: http://www.ninebynine.org/#Contact From Libby.Miller at bristol.ac.uk Thu Sep 16 18:04:44 2004 From: Libby.Miller at bristol.ac.uk (Libby Miller) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] Issue (?): URI of owl:Ontology header in the RDFS spec In-Reply-To: References: Message-ID: hia http://xmlns.com/foaf/0.1/ should do content negotiation so you get RDF if you ask for it, html if you don't. Unfortunately I think the stuff you get in rdf from that location has come adrift from the index.rdf version (which is the better version). This needs fixing, sorry. Libby On Thu, 16 Sep 2004, Richard Cyganiak wrote: > Hi everyone, > > I noticed another thing about the FOAF RDFS file. I'm not sure if I have a > point this time or if it's just nitpicking. > > The URL of the RDFS file is http://xmlns.com/foaf/0.1/index.rdf . > > The owl:Ontology header of this file says: > > > The two URIs don't match. When the RDFS file is processed, the processor > cannot know that the statements in the owl:Ontology element (dc:title, > dc:date etc.) are about the current file. The processor will believe that > they refer to another resource located somewhere else. > > The rdf:about URI should be changed, or an xml:base attribute added. Right? > > Thanks, > Richard > > > _______________________________________________ > rdfweb-dev mailing list > rdfweb-dev@vapours.rdfweb.org > wiki: http://rdfweb.org/topic/FoafProject > http://rdfweb.org/mailman/listinfo/rdfweb-dev > From richard at cyganiak.de Thu Sep 16 21:50:06 2004 From: richard at cyganiak.de (Richard Cyganiak) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] Issue (?): URI of owl:Ontology header in the RDFS spec In-Reply-To: References: Message-ID: <5F1B7674-082A-11D9-9A5B-000A95DABFDC@cyganiak.de> Hi Libby, > http://xmlns.com/foaf/0.1/ > > should do content negotiation so you get RDF if you ask for it, html > if you don't. Unfortunately I think the stuff you get in rdf from that > location has come adrift from the index.rdf version (which is the > better version). This needs fixing, sorry. I tried this now, and content negotiation works (Cool!) and the returned version is identical to the index.rdf one. So the only problem is that the URI of the owl:Ontology element does not match the document URL when you fetch from the index.rdf URL. This could be fixed by adding an xml:base="http://xmlns.com/foaf/0.1/", I think. I'll add it to the IssueTracker. Richard From bnowack at appmosphere.com Thu Sep 16 23:32:02 2004 From: bnowack at appmosphere.com (Benjamin Nowack) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] Issue (?): URI of owl:Ontology header in the RDFS spec In-Reply-To: <5F1B7674-082A-11D9-9A5B-000A95DABFDC@cyganiak.de> References: <5F1B7674-082A-11D9-9A5B-000A95DABFDC@cyganiak.de> Message-ID: On 16.09.2004 23:50:06, Richard Cyganiak wrote: > >I tried this now, and content negotiation works (Cool!) and the >returned version is identical to the index.rdf one. So the only problem >is that the URI of the owl:Ontology element does not match the document >URL when you fetch from the index.rdf URL. > >This could be fixed by adding an xml:base="http://xmlns.com/foaf/0.1/", >I think. Hm, I'm not sure if that needs "fixing"... I see the "index.rdf" doc as a copy/mirror of the the ontology doc located at "http://xmlns.com/foaf/0.1/" (via conneg). So the rdf/xml isn't about the "index.rdf" but the spec at the "official" URL. (i.e. you could put a copy on an arbitrary server, e.g. "http://example.com/foaf_spec.rdf" but would probably still keep the rdf:about="http://xmlns.com/foaf/0.1/" unless you wanted to talk about the copied spec.) As the foaf rdfs doc doesn't use rdf:ID for its terms, I don't think that an xml:base is needed. (it could perhaps be a (different) issue that the conneg "feature" leads to an overloaded URI/URL. We can't unambiguously use it to e.g. talk about the html representation vs the rdf/xml version at "http://xmlns.com/foaf/0.1/" ...) good night, benjamin -- Benjamin Nowack Kruppstr. 100 45145 Essen, Germany http://www.appmosphere.com/ > >I'll add it to the IssueTracker. > >Richard > > >_______________________________________________ >rdfweb-dev mailing list >rdfweb-dev@vapours.rdfweb.org >wiki: http://rdfweb.org/topic/FoafProject >http://rdfweb.org/mailman/listinfo/rdfweb-dev > > From saul at twenteenthcentury.com Fri Sep 17 03:09:52 2004 From: saul at twenteenthcentury.com (Saul Albert) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] distributed library project booklists in rdf Message-ID: <20040917030952.GG48246@chinabone.lth.bclub.org.uk> Dear RDFwebs, I've been lurking on this list for the *longest* time.. and I'm very aware of my technical inexperience.. much of it goes way over my head, so apologies if the question is misplaced or just silly: I've been working with a group on the distributed library project software - a books.burri.to style like thing but with less well sructured data, a miserable back end and basically a big pile of ugly hacks in php, but (unlike books.burri.to) people do actually use it to catalogue and share their books and metadata about stuff that otherwise might not be catalogued or accessible: - http://dlp.theps.net - http://dlp.theps.net/maps/ (consume.net style map of library nodes around limehouse town hall - http://twenteenthcentury.com/lth) I decided a while ago that it was beyond my short-term abilities to restructure the data model of the dlp and re-write it all from scratch, especially as people are actually using it. The value of the project then is simply data aggregation - at the moment it does not much with what is a very interesting data set (potentially this could be a way of generating (wikipedia style) the largest private library collection around given that there are now over 26 library nodes in 14 countries). Also, the prospect of semantic links between different-language materials through their cross-associations with other books / book lists etc. is an enticing one. So my question is this: If you were going to produce some interesting apps to scrape lots and lots of very interesting data sets form a bunch of distributed library nodes (basically, bibliographic information plus reviews, del.icio.us like tags (coming soon I hope) and geolocation / ownership / borrowing associations etc..) and do all kinds of magical associcative / locative / meaning-mangling stuff with it, what format would be most useful to you? I'll try to do standard library data exchange export formats for individual's librarys and the whole database of each node, but I'm interested to know if anyone has worked on any rdf that might be suitable for this.. I've seen this http://www.cellml.org/public/metadata/citations.html and jo did some graph models ages ago that would be useful if I could figure out what the foaf file they would come from would look like: http://dlpdev.theps.net/DlpGraphModels Basically I'm hoping someone will have done something similar - bibliographic records cross referenced with geolocation and other associations of ownership, tags etc... that I can just rip off and churn out from our databases.. anyone? thanks Saul. -- other people's sigs experience my ethereal equine art --- http://www.amandakoh.com and help to free art for all --- http://www.prodigalart.org We are all one. -- From stpeter at jabber.org Fri Sep 17 03:52:37 2004 From: stpeter at jabber.org (Peter Saint-Andre) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] full of questions Message-ID: After a hiatus of about a year, I've re-started serious research into integrating FOAF and other RDF vocabularies into Jabber/XMPP. I'm sure I'll have quite a few questions as I move along and work to write a Jabber Enhancement Proposal that defines the integration. Here are some starter questions: 1. Is it possible to define oneself as a member of a group? I like the fact that a Group is its own kind of Agent, but I think it would be helpful to be able to say "I'm a member of the group 'JSF Members' defined by http://www.jabber.org/members/foaf.rdf" (no FOAF file at that URL yet, but I'll make one soon). I realize that in true RDF fashion we'd expect the appropriate software agents to crawl around and find that group definition, then connect me with that group if its FOAF file says I am a member. Is there no shortcut like foaf:memberOf or something like that? 2. Connected to this: is there a property that enables me to say "I subscribe to this mailing list" or something of that kind? Is that simply an interest? We might not want to publish this Group, i.e., the list of folks who subscribe to a list (in fact as a list admin I always set access to the subscriber list to "list admin only" in Mailman), but I might want to identify myself as a subscriber to a list. I suppose making this an interest might work. 3. Are there any properties for things like marital status and other dating-related items (e.g., all the usual TLAs one sees in personals ads)? As you can imagine, such attributes are of more than passing interest in the IM world. 4. Did you folks ever settle any of the issues surrounding the representation of physical addresses? We're working to fully replace everything that currently exists in vCard, so address representation is necessary for us. 5. On the Jabber network, we might want to create FOAF representations for things like chatrooms, bots, and IM servers (currently many Jabber servers have their own vCards). We could see a chatroom as a kind of Group, but a bot is not a Person (though it is an Agent) and I would characterize a server as an AccountService of some kind (I don't see how it fits into any of the existing Agent categories). 6. An Internet-Draft defining the xmpp: URI scheme is pretty far along within the IETF, so it seems we might want to use that in representing the foaf:jabberID property. (Then again, we might not, given the wild escaping rules and such for URIs; a jabberID is at root a UTF-8 string that allows all sorts of nice internationalized support, whereas a URI needs to escape non-US-ASCII characters; so perhaps leaving things as they are would be better -- but at the least we need to specify whether a foaf:jabberID is an XMPP URI or a native XMPP address.) 7. Regarding the OnlineChatAccount property, I am wondering what would be appropriate for the accountServiceHomepage. As you may recall, the Jabber network has a distributed architecture, and anyone may run their own Jabber server (just as they may do with email). So it seems that each Jabber server would have its own homepage and that we can't define one generic URI here as might be able to do with Freenode (as in the example shown in the FOAF spec), ICQ, AIM, and so on. I'd like to help define the mapping between the abbreviated foaf:jabberID property and the full OnlineChatAccount representation, so we'd need to figure this out in order to make that happen. I'm happy to work on defining any of the attributes we're looking for that don't exist yet. More to come, I'm sure... Peter From Libby.Miller at bristol.ac.uk Fri Sep 17 04:56:48 2004 From: Libby.Miller at bristol.ac.uk (Libby Miller) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] Issue (?): URI of owl:Ontology header in the RDFS spec In-Reply-To: <5F1B7674-082A-11D9-9A5B-000A95DABFDC@cyganiak.de> References: <5F1B7674-082A-11D9-9A5B-000A95DABFDC@cyganiak.de> Message-ID: Thanks Richard! Libby On Thu, 16 Sep 2004, Richard Cyganiak wrote: > Hi Libby, > >> http://xmlns.com/foaf/0.1/ >> >> should do content negotiation so you get RDF if you ask for it, html if you >> don't. Unfortunately I think the stuff you get in rdf from that location >> has come adrift from the index.rdf version (which is the better version). >> This needs fixing, sorry. > > I tried this now, and content negotiation works (Cool!) and the returned > version is identical to the index.rdf one. So the only problem is that the > URI of the owl:Ontology element does not match the document URL when you > fetch from the index.rdf URL. > > This could be fixed by adding an xml:base="http://xmlns.com/foaf/0.1/", I > think. > > I'll add it to the IssueTracker. > > Richard > > From danbri at w3.org Fri Sep 17 08:35:46 2004 From: danbri at w3.org (Dan Brickley) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] Issue (?): URI of owl:Ontology header in the RDFS spec In-Reply-To: References: <5F1B7674-082A-11D9-9A5B-000A95DABFDC@cyganiak.de> Message-ID: <20040917083546.GA7435@homer.w3.org> * Benjamin Nowack [2004-09-17 01:32+0200] > On 16.09.2004 23:50:06, Richard Cyganiak wrote: > > > >I tried this now, and content negotiation works (Cool!) and the > >returned version is identical to the index.rdf one. So the only problem > >is that the URI of the owl:Ontology element does not match the document > >URL when you fetch from the index.rdf URL. > > > >This could be fixed by adding an xml:base="http://xmlns.com/foaf/0.1/", > >I think. > Hm, I'm not sure if that needs "fixing"... > I see the "index.rdf" doc as a copy/mirror of the the ontology > doc located at "http://xmlns.com/foaf/0.1/" (via conneg). So the rdf/xml > isn't about the "index.rdf" but the spec at the "official" URL. > (i.e. you could put a copy on an arbitrary server, e.g. > "http://example.com/foaf_spec.rdf" but would probably still keep the > rdf:about="http://xmlns.com/foaf/0.1/" unless you wanted to talk about > the copied spec.) > As the foaf rdfs doc doesn't use rdf:ID for its terms, I don't think > that an xml:base is needed. That's how I see it too. The index.rdf is a convenience resource. There's also a copy embedded raw in the xhtml version, as well as content negotiable. A 3rd way to see it was as entity-escaped colourised markup (generated with XSLT), but I removed that as the generator script was broken and leaving the human-visible markup appendix unreadable. It also made the spec really long. The markup was designed so it'd generated the same triples. Also btw should work so that archival copies of the namespace could be made, eg. 20040901-index.rdf or whatever, and generate triples that provide a 'historical' view that talks about the same things, rather than talking about properties/classes whose URIs contain '20040901' etc... For all that, it should be harmless to include the xml:base declaration, might make the intent clearer. Dan > > (it could perhaps be a (different) issue that the conneg "feature" > leads to an overloaded URI/URL. We can't unambiguously use it to e.g. > talk about the html representation vs the rdf/xml version at > "http://xmlns.com/foaf/0.1/" ...) > > good night, > benjamin > > -- > Benjamin Nowack > > Kruppstr. 100 > 45145 Essen, Germany > http://www.appmosphere.com/ > > > > >I'll add it to the IssueTracker. > > > >Richard > > > > > >_______________________________________________ > >rdfweb-dev mailing list > >rdfweb-dev@vapours.rdfweb.org > >wiki: http://rdfweb.org/topic/FoafProject > >http://rdfweb.org/mailman/listinfo/rdfweb-dev > > > > > > > _______________________________________________ > rdfweb-dev mailing list > rdfweb-dev@vapours.rdfweb.org > wiki: http://rdfweb.org/topic/FoafProject > http://rdfweb.org/mailman/listinfo/rdfweb-dev From rich.boakes at port.ac.uk Fri Sep 17 09:11:06 2004 From: rich.boakes at port.ac.uk (rich boakes) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] Issue (?): URI of owl:Ontology header in the RDFS spec In-Reply-To: <20040917083546.GA7435@homer.w3.org> References: <5F1B7674-082A-11D9-9A5B-000A95DABFDC@cyganiak.de> <20040917083546.GA7435@homer.w3.org> Message-ID: <414AAA2A.5050009@port.ac.uk> danbri wrote: > A 3rd way to see it was as entity-escaped colourised > markup (generated with XSLT), but I removed that as > the generator script was broken ... He Dan, Does the XSLT part still exist and was it at a level where it might be generally usable? Rich From danbri at w3.org Fri Sep 17 09:19:30 2004 From: danbri at w3.org (Dan Brickley) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] Issue (?): URI of owl:Ontology header in the RDFS spec In-Reply-To: <414AAA2A.5050009@port.ac.uk> References: <5F1B7674-082A-11D9-9A5B-000A95DABFDC@cyganiak.de> <20040917083546.GA7435@homer.w3.org> <414AAA2A.5050009@port.ac.uk> Message-ID: <20040917091930.GI7435@homer.w3.org> * rich boakes [2004-09-17 10:11+0100] > danbri wrote: > > A 3rd way to see it was as entity-escaped colourised > > markup (generated with XSLT), but I removed that as > > the generator script was broken ... > > He Dan, > > Does the XSLT part still exist and was it at a level where > it might be generally usable? The XSLT is from Max Froumentin, see http://www.w3.org/People/maxf/ under "XML file pretty printer" http://www.w3.org/2003/02/colour-xml-serializer.xsl I think it needs a few args for setting the colours, and can't in XSLT 1 be entirely general purpose. But the code's up there... cheers, Dan From richard at cyganiak.de Fri Sep 17 10:58:03 2004 From: richard at cyganiak.de (Richard Cyganiak) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] Issue (?): URI of owl:Ontology header in the RDFS spec In-Reply-To: <20040917083546.GA7435@homer.w3.org> References: <5F1B7674-082A-11D9-9A5B-000A95DABFDC@cyganiak.de> <20040917083546.GA7435@homer.w3.org> Message-ID: <72D47006-0898-11D9-9A5B-000A95DABFDC@cyganiak.de> Hi Dan, hi Benjamin, thanks for your answers. Am 17.09.2004 um 10:35 schrieb Dan Brickley: > * Benjamin Nowack [2004-09-17 01:32+0200] >> On 16.09.2004 23:50:06, Richard Cyganiak wrote: >>> the URI of the owl:Ontology element does not match the document >>> URL when you fetch from the index.rdf URL. >>> >>> This could be fixed by adding an >>> xml:base="http://xmlns.com/foaf/0.1/", >>> I think. >> Hm, I'm not sure if that needs "fixing"... > That's how I see it too. The index.rdf is a convenience resource. Here is why I think the document currently served from .../index.rdf is broken: Imagine I want to open the FOAF spec in my RDFS/OWL ontology editor. So I google for "foaf rdfs" or search the HTML spec for a link to the RDFS and end up with the .../index.rdf link. I open it in my editor. All the classes and properties show up just fine. But there's no metadata about the ontology itself, like title, date and so on, like you have in pretty much every RDFS file you can find on the net. So they didn't include that in the FOAF RDFS? How lazy! On closer inspection I see that the info is there in the file, but my ontology editor didn't show it because it was unable to associate the owl:Ontology header with the current document. The editor "thought" that, for some reason, the file contained metadata about some other ontology somewhere on the Web with no apparent relation to the current file. Needs fixing or not, from that point of view? I absolutely agree with all your good points about index.rdf just being a mirror, archival copies, and having the same triples in all versions. None of these would be harmed by an xml:base declaration. Benjamin: >> As the foaf rdfs doc doesn't use rdf:ID for its terms, I don't think >> that an xml:base is needed. As I see it, xml:base is used for three things: 1. building URIs from rdf:ID, 2. building absolute URIs from relative URIs, as in rdf:about="#foo". 3. determining what the URI of the document itself is, as in rdf:about="". 1 and 2 are not relevant for the FOAF RDFS. 3 is. Dan: > For all that, it should be harmless to include the xml:base > declaration, > might make the intent clearer. That would be great! Thanks, Richard From danbri at w3.org Fri Sep 17 11:09:06 2004 From: danbri at w3.org (Dan Brickley) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] Issue (?): URI of owl:Ontology header in the RDFS spec In-Reply-To: <72D47006-0898-11D9-9A5B-000A95DABFDC@cyganiak.de> References: <5F1B7674-082A-11D9-9A5B-000A95DABFDC@cyganiak.de> <20040917083546.GA7435@homer.w3.org> <72D47006-0898-11D9-9A5B-000A95DABFDC@cyganiak.de> Message-ID: <20040917110906.GM7435@homer.w3.org> * Richard Cyganiak [2004-09-17 12:58+0200] > Hi Dan, hi Benjamin, > > thanks for your answers. > > Am 17.09.2004 um 10:35 schrieb Dan Brickley: > >* Benjamin Nowack [2004-09-17 01:32+0200] > >>On 16.09.2004 23:50:06, Richard Cyganiak wrote: > >>>the URI of the owl:Ontology element does not match the document > >>>URL when you fetch from the index.rdf URL. > >>> > >>>This could be fixed by adding an > >>>xml:base="http://xmlns.com/foaf/0.1/", > >>>I think. > >>Hm, I'm not sure if that needs "fixing"... > > >That's how I see it too. The index.rdf is a convenience resource. > > > Here is why I think the document currently served from .../index.rdf is > broken: > > Imagine I want to open the FOAF spec in my RDFS/OWL ontology editor. So > I google for "foaf rdfs" or search the HTML spec for a link to the RDFS > and end up with the .../index.rdf link. I open it in my editor. All the > classes and properties show up just fine. But there's no metadata about > the ontology itself, like title, date and so on, like you have in > pretty much every RDFS file you can find on the net. So they didn't > include that in the FOAF RDFS? How lazy! > > On closer inspection I see that the info is there in the file, but my > ontology editor didn't show it because it was unable to associate the > owl:Ontology header with the current document. The editor "thought" > that, for some reason, the file contained metadata about some other > ontology somewhere on the Web with no apparent relation to the current > file. > > Needs fixing or not, from that point of view? I'm not familiar enough with the behaviour of the various ontology editing tools out there, but it does sound like an xml:base would help them out. And it should be harmless for other purposes, so I'll add it. In the Semantic Web Best Practices WG at W3C we are just starting to examine such issues within a new 'Vocabulary Management' Task Force. I'm hoping FOAF can be made into a case study area there, so I expect to come back to this issue later, and see how other namespaces are managing this problem. In meantime, I'll add xml:base as part of next update to the site. cheerrs, Dan > I absolutely agree with all your good points about index.rdf just being > a mirror, archival copies, and having the same triples in all versions. > None of these would be harmed by an xml:base declaration. > > > Benjamin: > >>As the foaf rdfs doc doesn't use rdf:ID for its terms, I don't think > >>that an xml:base is needed. > > As I see it, xml:base is used for three things: > > 1. building URIs from rdf:ID, > 2. building absolute URIs from relative URIs, as in rdf:about="#foo". > 3. determining what the URI of the document itself is, as in > rdf:about="". > > 1 and 2 are not relevant for the FOAF RDFS. 3 is. > > > Dan: > >For all that, it should be harmless to include the xml:base > >declaration, > >might make the intent clearer. > > That would be great! > > Thanks, > Richard From bnowack at appmosphere.com Fri Sep 17 12:12:52 2004 From: bnowack at appmosphere.com (Benjamin Nowack) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] Issue (?): URI of owl:Ontology header in the RDFS spec In-Reply-To: <72D47006-0898-11D9-9A5B-000A95DABFDC@cyganiak.de> References: <5F1B7674-082A-11D9-9A5B-000A95DABFDC@cyganiak.de> <20040917083546.GA7435@homer.w3.org> <72D47006-0898-11D9-9A5B-000A95DABFDC@cyganiak.de> Message-ID: On 17.09.2004 12:58:03, Richard Cyganiak wrote: >Here is why I think the document currently served from .../index.rdf is >broken: > >Imagine I want to open the FOAF spec in my RDFS/OWL ontology editor. So >I google for "foaf rdfs" or search the HTML spec for a link to the RDFS >and end up with the .../index.rdf link. I open it in my editor. All the >classes and properties show up just fine. But there's no metadata about >the ontology itself, like title, date and so on, like you have in >pretty much every RDFS file you can find on the net. So they didn't >include that in the FOAF RDFS? How lazy! > >On closer inspection I see that the info is there in the file, but my >ontology editor didn't show it because it was unable to associate the >owl:Ontology header with the current document. The editor "thought" >that, for some reason, the file contained metadata about some other >ontology somewhere on the Web with no apparent relation to the current >file. > >Needs fixing or not, from that point of view? if your ontology editor really behaves that way, then I guess it'd need fixing, too ;) afaik, none of the resources described in the foaf spec are related to the "index.rdf" url. all the terms are identified by *absolute* uris, e.g. rdf:about="http://xmlns.com/foaf/0.1/Person". And so is the ontology. Why should your editor read the terms but not the ontology resource? And adding an xml:base="http://xmlns.com/foaf/0.1/" wouldn't actually change the ontology's uri as it doesn't have a relative uri assigned. either version leads to exactly the same triples. But I agree with Dan, adding the base doesn't hurt. >I absolutely agree with all your good points about index.rdf just being >a mirror, archival copies, and having the same triples in all versions. >None of these would be harmed by an xml:base declaration. see above, true. see below, maybe false. >Benjamin: >>> As the foaf rdfs doc doesn't use rdf:ID for its terms, I don't think >>> that an xml:base is needed. > >As I see it, xml:base is used for three things: > >1. building URIs from rdf:ID, >2. building absolute URIs from relative URIs, as in rdf:about="#foo". >3. determining what the URI of the document itself is, as in >rdf:about="". > >1 and 2 are not relevant for the FOAF RDFS. 3 is. but there is no rdf:about="" in either of "/" or "index.rdf" docs, or am I looking at the wrong docs? In the latter case I absolutely agree with you, but I get both times here. And to be even more nitpicking (sorry ;), *if* we add an xml:base to the spec doc at index.rdf, we couldn't easily add an additional which would make it possible to describe circulated copies. Imagine we had a tool that created a copy of the spec every month to keep those "snapshots" dan mentioned in the wiki. It would be easy to add a dc:date to the additional ontology header and then save this doc whereever you want. the dc:date would always point to the specific copy. (well, this doesn't help with the terms' absolute uris, if those copies were published, but still..) ;) benjamin -- Benjamin Nowack Kruppstr. 100 45145 Essen, Germany http://www.appmosphere.com/ From stpeter at jabber.org Fri Sep 17 15:33:37 2004 From: stpeter at jabber.org (Peter Saint-Andre) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] QuoteIssue Message-ID: I have created a QuoteIssue in the IssueTracker: http://rdfweb.org/topic/QuoteIssue Abstract: The recent Pew Internet survey on instant messaging noted that the most popular form of content added to IM profile pages (think the ICQ website) is "funny quotes or other sayings" that the person identifies with in some way. The numbers: 42% of profile creators included quotes, whereas only 18% included links to websites of interest (think foaf:interest). So it seems that it would be good for FOAF to include a foaf:quote property of some kind. Let me know if more detailed information is desired, and do tell me how to proceed with this. Shall I write up some proposed text and examples? I am happy to be the issue owner for this. Peter From stpeter at jabber.org Fri Sep 17 15:34:42 2004 From: stpeter at jabber.org (Peter Saint-Andre) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] AccountServiceIssue Message-ID: I have created an AccountServiceIssue in the IssueTracker: http://rdfweb.org/topic/AccountServiceIssue Abstract: On the Jabber network, we might want to create FOAF representations for things like chatrooms, bots, and IM servers. In particular, currently many Jabber servers have their own vCards and we would want to replace those with a FOAF representation. I think that an IM server is a kind of Agent that is not yet discussed elsewhere, which I would characterize as an AccountService. This terms needs further definition but I think it captures the core idea (related, naturally, to the existing accountServiceHomepage property). Let me know if more detailed information is desired, and do tell me how to proceed with this. Shall I write up some proposed text and examples? I am happy to be the issue owner for this. Peter From richard at cyganiak.de Fri Sep 17 18:26:29 2004 From: richard at cyganiak.de (Richard Cyganiak) Date: Mon Dec 18 17:23:57 2006 Subject: OT: base URIs and semantics (was: Re: [rdfweb-dev] Issue (?): URI of owl:Ontology header in the RDFS spec) In-Reply-To: References: <5F1B7674-082A-11D9-9A5B-000A95DABFDC@cyganiak.de> <20040917083546.GA7435@homer.w3.org> <72D47006-0898-11D9-9A5B-000A95DABFDC@cyganiak.de> Message-ID: <17A7E7E6-08D7-11D9-9A5B-000A95DABFDC@cyganiak.de> Hi Benjamin, to keep this short: I believe that a document's base URI is part of its semantics. Change the base URI, change the semantics. This is not stated in any spec, but it's a way of thinking that works quite well for me. Triples are not everything. There are containers for triples too. rdf:about="" and rdf:about="$baseURI" are ways to say things about containers for triples. Remember this when moving these containers around. xml:base is a way to make sure you can move these containers around without breaking any statements about them. But, granted, this is one of those "Here be dragons" areas of RDF semantics, so you are free to disagree, and I can't prove wrong anything you said below. Richard > On 17.09.2004 12:58:03, Richard Cyganiak wrote: >> Here is why I think the document currently served from .../index.rdf >> is >> broken: >> >> Imagine I want to open the FOAF spec in my RDFS/OWL ontology editor. >> So >> I google for "foaf rdfs" or search the HTML spec for a link to the >> RDFS >> and end up with the .../index.rdf link. I open it in my editor. All >> the >> classes and properties show up just fine. But there's no metadata >> about >> the ontology itself, like title, date and so on, like you have in >> pretty much every RDFS file you can find on the net. So they didn't >> include that in the FOAF RDFS? How lazy! >> >> On closer inspection I see that the info is there in the file, but my >> ontology editor didn't show it because it was unable to associate the >> owl:Ontology header with the current document. The editor "thought" >> that, for some reason, the file contained metadata about some other >> ontology somewhere on the Web with no apparent relation to the current >> file. >> >> Needs fixing or not, from that point of view? > if your ontology editor really behaves that way, then I guess it'd need > fixing, too ;) > afaik, none of the resources described in the foaf spec are related > to the "index.rdf" url. all the terms are identified by *absolute* > uris, > e.g. rdf:about="http://xmlns.com/foaf/0.1/Person". And so is the > ontology. > Why should your editor read the terms but not the ontology resource? > And > adding an xml:base="http://xmlns.com/foaf/0.1/" wouldn't actually > change > the ontology's uri as it doesn't have a relative uri assigned. either > version leads to exactly the same triples. But I agree with Dan, adding > the base doesn't hurt. > >> I absolutely agree with all your good points about index.rdf just >> being >> a mirror, archival copies, and having the same triples in all >> versions. >> None of these would be harmed by an xml:base declaration. > see above, true. see below, maybe false. > > >> Benjamin: >>>> As the foaf rdfs doc doesn't use rdf:ID for its terms, I don't think >>>> that an xml:base is needed. >> >> As I see it, xml:base is used for three things: >> >> 1. building URIs from rdf:ID, >> 2. building absolute URIs from relative URIs, as in rdf:about="#foo". >> 3. determining what the URI of the document itself is, as in >> rdf:about="". >> >> 1 and 2 are not relevant for the FOAF RDFS. 3 is. > but there is no rdf:about="" in either of "/" or "index.rdf" docs, or > am I looking at the wrong docs? In the latter case I absolutely agree > with you, but I get > > both times here. > > And to be even more nitpicking (sorry ;), *if* we add an xml:base to > the spec doc at index.rdf, we couldn't easily add an additional > which would make it possible to describe > circulated copies. Imagine we had a tool that created a copy of the > spec every month to keep those "snapshots" dan mentioned in the wiki. > It would be easy to add a dc:date to the additional ontology header > and then save this doc whereever you want. the dc:date would always > point to the specific copy. (well, this doesn't help with the terms' > absolute uris, if those copies were published, but still..) > > > ;) > benjamin > > -- > Benjamin Nowack > > Kruppstr. 100 > 45145 Essen, Germany > http://www.appmosphere.com/ > > From danny.ayers at gmail.com Fri Sep 17 20:11:33 2004 From: danny.ayers at gmail.com (Danny Ayers) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] Re: RDF model sync, HTTP deltas, syndicated feeds (and a bounty) In-Reply-To: References: <1f2ed5cd0409130426540c86f9@mail.gmail.com> Message-ID: <1f2ed5cd0409171311669fec64@mail.gmail.com> (thanks stpeter) Update - an entry-oriented diff plugin for Apache has been put together - mod_speedyfeed, see: http://asdf.blogs.com/asdf/2004/09/mod_speedyfeed__1.html some discussion: http://www.intertwingly.net/blog/2004/09/17/mod-speedyfeed I've asked whether it might be suitable for fairly arbitrary RDF/XML - aside from simplifying sync/sharing big graphs, this would be a boon to the presence foafsters, so suggestions appreciated: http://dannyayers.com/archives/2004/09/17/mod_speedyfeed-for-rdfxml/ Cheers, Danny. On Mon, 13 Sep 2004 15:50:11 -0600, Peter Saint-Andre wrote: > In article <1f2ed5cd0409130426540c86f9@mail.gmail.com>, > Danny Ayers wrote: > > > http://bobwyman.pubsub.com/main/2004/09/using_rfc3229_w.html > > Bob also mentions that a true pubsub approach would be best in the long > term; for instance, see: > > http://www.ietf.org/internet-drafts/draft-saintandre-atompub-notify-01.tx > t > > /psa > > _______________________________________________ > rdfweb-dev mailing list > rdfweb-dev@vapours.rdfweb.org > wiki: http://rdfweb.org/topic/FoafProject > http://rdfweb.org/mailman/listinfo/rdfweb-dev > -- http://dannyayers.com From zednenem at psualum.com Sat Sep 18 03:55:14 2004 From: zednenem at psualum.com (David Menendez) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] Issue (?): URI of owl:Ontology header in the RDFS spec In-Reply-To: <72D47006-0898-11D9-9A5B-000A95DABFDC@cyganiak.de> Message-ID: Richard Cyganiak writes: > Here is why I think the document currently served from .../index.rdf > is broken: > > Imagine I want to open the FOAF spec in my RDFS/OWL ontology editor. > So I google for "foaf rdfs" or search the HTML spec for a link to the > RDFS and end up with the .../index.rdf link. I open it in my editor. > All the classes and properties show up just fine. But there's no > metadata about the ontology itself, like title, date and so on, like > you have in pretty much every RDFS file you can find on the net. So > they didn't include that in the FOAF RDFS? How lazy! > > On closer inspection I see that the info is there in the file, but my > ontology editor didn't show it because it was unable to associate the > owl:Ontology header with the current document. The editor "thought" > that, for some reason, the file contained metadata about some other > ontology somewhere on the Web with no apparent relation to the > current file. > > Needs fixing or not, from that point of view? To me, this sounds more like a flaw in the ontology editor. Specifically, it isn't honoring rdfs:isDefinedBy. The FOAF schema has a structure like this (in Turtle): @prefix foaf: . a owl:Ontology. foaf:Agent a owl:Class ; rdfs:isDefinedBy . foaf:Person a owl:Class ; rdfs:isDefinedBy . and so forth. As I see it, the proper behavior for an ontology editor seeing this would be: 1. Find all resources which are instances of owl:Ontology 2. For each ontology, find all resources related to that ontology by rdfs:isDefinedBy. For example, my schema for TDL 3 [1] describes how some of the new terms are related to obsolete terms from earlier versions. An ontology editor would use the rdfs:isDefinedBy relations to learn that tdl:mentions is part of TDL 3, while tdl:commentsOn is not. [1] -- David Menendez | "In this house, we obey the laws | of thermodynamics!" From richard at cyganiak.de Sat Sep 18 11:34:56 2004 From: richard at cyganiak.de (Richard Cyganiak) Date: Mon Dec 18 17:23:57 2006 Subject: OT: rdfs:isDefinedBy and ontology editors (was: Re: [rdfweb-dev] Issue (?): URI of owl:Ontology header in the RDFS spec) In-Reply-To: References: Message-ID: Hi David, Am 18.09.2004 um 05:55 schrieb David Menendez: > To me, this sounds more like a flaw in the ontology editor. > Specifically, it isn't honoring rdfs:isDefinedBy. The semantics of rdfs:isDefinedBy are fuzzy. People might use it for what you suggest, but a couple of different uses would also be covered by the loose definition in the spec. Basing the behaviour of a generic tool on that doesn't seem like a good idea to me. > As I see it, the proper behavior for an ontology editor seeing this > would be: > > 1. Find all resources which are instances of owl:Ontology > 2. For each ontology, find all resources related to that > ontology by rdfs:isDefinedBy. That doesn't work in the common case since both the owl:Ontology header and rdfs:isDefinedBy are optional. > [1] Nice style for the semicolons! Surely makes editing easier. Richard From danny.ayers at gmail.com Sat Sep 18 14:27:03 2004 From: danny.ayers at gmail.com (Danny Ayers) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] TDL and IBIS Message-ID: <1f2ed5cd0409180727197318a2@mail.gmail.com> Hi David, At some point in the not-too-distant future I plan to update the RDF vocabulary [1] I did for IBIS to be more OWLish (IBIS = Issue-Based Information Systems, a collaborative problem analysis/solving technique with a threaded-message model). There's considerable overlap with TDL, so I think it would be worth including equivalentProperty/subPropertyOf etc mappings as appropriate. If and when you have a minute I'd be grateful if you could look over the current schema and see what you reckon. Cheers, Danny. [1] http://purl.org/ibis -- http://dannyayers.com From bnowack at appmosphere.com Tue Sep 21 18:36:52 2004 From: bnowack at appmosphere.com (Benjamin Nowack) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] agenda item suggestion for wednesday's chat Message-ID: hi FOAFers, is the chat going to happen tomorrow (5:30 BST, IIRC)? I'm currently trying to implement some tools related to the FoafCommunityProcess[1]. If the chat is taking place, I'd like to discuss some items of the "Things To Aspire To" section: "identification of term maturity" the spec currently uses "unstable", "testing", "stable" which can be used to describe the current stage of a term's life cycle. If a term is planned to be removed from the term set we could use OWL's deprecation constructs to flag a property or class as deprecated. I wonder if there could be (corner ?) cases where a term's status went from "unstable" to "testing" but after that didn't really get deployed. If the term is not planned to be deprecated, is there a need/way to indicate that this term "was tested, will probably not move to stable but is going to be kept in the spec". This is related to last week's "which terms for names" discussion. Some terms might be in the spec but their use is not recommended any more. Or put in other words: "Tool developers, use these terms. they'll move to 'stable' soon". perhaps this discussion can be replaced by thinking about adding "last modified" or "status changed" annotations to each term. "term documentation" I volunteered at foafcamp to build some kind of a "term browser" for easier access to the description of selected foaf terms. Each term view could show a term's standard information (label, comment, term_status, relations, description, rdf/xml), but also more detailed stuff such as examples, links to tools, use cases, best practices, similar/related terms, version info, issues, todos, etc. Any thoughts on useful documentation types? Just some ideas.. cheers, benjamin [1] http://rdfweb.org/topic/FOAFCommunityProcess -- Benjamin Nowack Kruppstr. 100 45145 Essen, Germany http://www.appmosphere.com/ From tony-rdf at kasei.com Tue Sep 21 22:42:30 2004 From: tony-rdf at kasei.com (Tony Bowden) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] foaf:dateOfBirth In-Reply-To: References: <479B5AE2-E3C7-11D8-9F63-000D932A33F0@atommic.com> Message-ID: <20040921224230.GP8052@soto.kasei.com> On Sun, Aug 01, 2004 at 03:46:31PM +0100, Libby Miller wrote: > You could also use the bio vocab: > [[ > > > 1970-06-15 > Brentwood, Essex, United Kingdom > > > ]] > http://vocab.org/bio/0.1/#Birth Are there other vocabs that include a wider range of events? Rather than just saying that I foaf:knows someone, I'd often prefer to say that I met them at a certain event. What's the preferred way to do something like this? Tony From kjetil at kjernsmo.net Tue Sep 21 23:06:38 2004 From: kjetil at kjernsmo.net (Kjetil Kjernsmo) Date: Mon Dec 18 17:23:57 2006 Subject: Inference about the nature of relationships (was Re: [rdfweb-dev] foaf:dateOfBirth) In-Reply-To: <20040921224230.GP8052@soto.kasei.com> References: <479B5AE2-E3C7-11D8-9F63-000D932A33F0@atommic.com> <20040921224230.GP8052@soto.kasei.com> Message-ID: <200409220106.38501.kjetil@kjernsmo.net> On onsdag 22. september 2004, 00:42, Tony Bowden wrote: > Are there other vocabs that include a wider range of events? Rather > than just saying that I foaf:knows someone, I'd often prefer to say > that I met them at a certain event. What's the preferred way to do > something like this? Eh, well, I had a longish discussion with someone about something like this. Since you brought it up, and I feel I lack understanding, but nevertheless feel that I can elaborate on the problems and on the lack of understanding, I'll follow up anyway. Hope it doesn't sound too stupid to the gurus... :-) Here's my idea: I would like to not only say that I know someone, but why I do know them. Say, for example, about half of the people in my addressbook, I know because they, like me, participate in the sport of orienteering. Furthermore, in my addressbook, I have categorised them as orienteers, and I'd like to express that somehow in my FOAF. Now, one may advocate that just mark yourself up as ed in orienteering. One might then guess that if two people know each other and both share a common interest, they know each other because of that common interest. Well, I have an example where that inference would be wrong: My favorite is actually the even more exotic sport of ski-orienteering, where there are only about 1000 participants in the world. What's the chance of someone sharing interests not knowing eachother because of that...? :-) Well, the other day, it turned out that the guy making a Debian package of GRASS 5.7 is a ski-orienteer. And since I spent less than five hours working with his packages, the chance that I will markup that anytime soon is little, but my interest in ski-orienteering is something that I'll markup as soon as I come up with a good URI for it... I figured maybe the relationship vocab would do this: http://purl.org/vocab/relationship/ particularly the participant(In) properties...? Now, our discussion ended when it was noted that FOAF describes relationships in the form of arcs, and so you have to describe the arc somehow...? Cheers, Kjetil -- Kjetil Kjernsmo Astrophysicist/IT Consultant/Skeptic/Ski-orienteer/Orienteer/Mountaineer kjetil@kjernsmo.net webmaster@skepsis.no editor@learn-orienteering.org Homepage: http://www.kjetil.kjernsmo.net/ OpenPGP KeyID: 6A6A0BBC From danbri at w3.org Wed Sep 22 09:36:47 2004 From: danbri at w3.org (Dan Brickley) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] [collab@sims] CFP: Beyond Personalization 2005 Message-ID: <20040922093647.GD20521@homer.w3.org> It'd be interesting to see some RDF, FOAF and SemWeb apps... ----- Forwarded message from "Sean M. McNee" ----- From: "Sean M. McNee" Date: Tue, 21 Sep 2004 16:10:02 -0500 To: collab@sims.berkeley.edu Subject: [collab@sims] CFP: Beyond Personalization 2005 Message-Id: Call for Papers (with apologies for cross-posting) ------------------------------------------------------------ Beyond Personalization 2005 A Workshop on the Next Stage of Recommender Systems Research ------------------------------------------------------------ http://www.cs.umn.edu/Research/GroupLens/beyond2005/ ------------------------------------------------------------ In conjunction with the 2005 International Conference on Intelligent User Interfaces (IUI 2005) San Diego, California January 9-12, 2005 http://www.iuiconf.org ------------------------- Workshop Topics and Goals ------------------------- This workshop intends to bring recommender systems researchers and practitioners together in order to discuss the current state of recommender systems research, both on existing and emerging research topics, and to determine how research in this area should proceed. We are at a pivotal point in recommender systems research where researchers are both looking inward at what recommender systems are and looking outward at where recommender systems can be applied, and the implications of applying them out 'in the wild.' This creates a unique opportunity to both reassess the current state of research and directions research is taking in the near and long term. This workshop will focus on the following four main topics: 1. Understanding and trusting recommender systems. Do users understand and trust the recommendations they receive from recommender systems, what kinds of information do recommenders need to provide to users to build trust, and how difficult is it to regain trust in a recommender if it is lost? 2. User interfaces for recommender systems. What are good ways to present recommendations to users, how do you integrate recommenders into the displays of existing information systems, and how can interfaces encourage users to provide ratings in order to 'close the loop' for recommendations, that is, how can you get users to consume the items recommended and then tell the system how good the recommendations are? 3. The future of recommendation algorithms and their metrics. How can we generate better individual and group recommendations, develop new metrics and evaluation criteria for recommendations, and achieve cross-domain recommendations? 4. Social consequences and opportunities of recommender systems. How do individuals and groups of people respond to recommendations, how can recommendations be integrated with online and real world communities, and in what ways do recommendations affect social organizations? ----------------- Intended Audience ----------------- The workshop is intended for both established researchers and practioners in the domain of recommender systems as well as for new researchers and students with interesting ideas on recommender systems and their future. Participants do not have to come from a specific application domain, as long as their research or ideas are on one of the main topics of the workshop. ---------------- Paper Submission ---------------- Two types of contributions are invited: 1. Papers describing (ongoing) work on one or more of the topics for the workshop. 2. Position statements regarding one or more of the topics for the workshop. Both papers and position statements should be prepared according to the IUI 2005 Instructions for Authors, and should not exceed 6 pages for papers and 2 pages for position statements. Acceptable formats are PostScript and PDF. Please send the paper or position statement not later than 8 November 2004 by e-mail to mcnee@cs.umn.edu, cc: Mark.vanSetten@telin.nl Each paper and position statement will be reviewed by at least two reviewers. Accepted contributions will be published in the workshop proceedings and will be available on the Web before the workshop begins. Authors whose papers or position statements are accepted to this workshop are expected to attend both the workshop and IUI 2005. The workshop format will be panel-based with the authors of the best submitted papers appearing on the panels and include interactive sessions with all participants. --------------- Important Dates --------------- 8 Nov 2004: Paper Submission Deadline 28 Nov 2004: Paper Acceptance Notification 6 Dec 2004: Camera-ready Copies Due (and last day of Early Registration for IUI 2005) 10 Dec 2004: Papers Available from Website 9 Jan 2005: Workshop at IUI 2005 --------------- Workshop Chairs --------------- Mark van Setten Telematica Instituut P.O. Box 589 7500 AN Enschede The Netherlands E-mail: Mark.vanSetten@telin.nl Sean M. McNee GroupLens Research Department of Computer Science and Engineering University of Minnesota Minneapolis, MN, 55455 USA E-mail: mcnee@cs.umn.edu Joseph A. Konstan GroupLens Research Department of Computer Science and Engineering University of Minnesota Minneapolis, MN, 55455 USA E-mail: konstan@cs.umn.edu ----------------- Program Committee ----------------- Loren Terveen - University of Minnesota (USA) Liliana Ardissono - University of Torino (Italy) Jon Herlocker - Oregon State University (USA) Barry Smyth - University College Dublin and Changing Worlds (Ireland) Anton Nijholt - University of Twente (The Netherlands) --------- Posted to the collab@sims.berkeley.edu mailing list. To unsubscribe, send an email message to Majordomo@sims.berkeley.edu with the phrase "unsubscribe collab" in the body of the message. ----- End forwarded message ----- From danbri at w3.org Wed Sep 22 14:23:59 2004 From: danbri at w3.org (Dan Brickley) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] agenda item suggestion for wednesday's chat In-Reply-To: References: Message-ID: <20040922142359.GE5627@homer.w3.org> Hi Benjamin, all, * Benjamin Nowack [2004-09-21 20:36+0200] > > hi FOAFers, > > is the chat going to happen tomorrow (5:30 BST, IIRC)? Yes, though I didn't remember to announce it! Ahem... so that's in 3 hours time, for anyone who wants to stop by. #foaf IRC channel, irc.freenode.net aka irc://irc.freenode.net/foaf It should be pretty informal, given both the lack of notice and the fact I'm distracted with other things... > I'm currently trying to implement some tools related to > the FoafCommunityProcess[1]. If the chat is taking place, > I'd like to discuss some items of the "Things To Aspire > To" section: > > "identification of term maturity" > the spec currently uses "unstable", "testing", "stable" > which can be used to describe the current stage of a > term's life cycle. If a term is planned to be removed > from the term set we could use OWL's deprecation > constructs to flag a property or class as deprecated. > > I wonder if there could be (corner ?) cases where a > term's status went from "unstable" to "testing" but > after that didn't really get deployed. If the term > is not planned to be deprecated, is there a need/way > to indicate that this term "was tested, will probably > not move to stable but is going to be kept in the > spec". I think that's a case worth documenting in prose at least. Not clear how long it takes before we realise a term isn't going to end up in very widespread use. Also it might be argued that a few small-scale uses can still be of value. For eg., the stability vocab itself has some potential, but is unlikely to ever be used more than ~1000 or so times, because it is of fairly narrow applicability. > This is related to last week's "which terms > for names" discussion. Some terms might be in the > spec but their use is not recommended any more. Or > put in other words: "Tool developers, use these > terms. they'll move to 'stable' soon". perhaps this > discussion can be replaced by thinking about adding > "last modified" or "status changed" annotations to > each term. > There is some work in the Dublin Core Usage Board on this too. I'm hoping to attend their next f2f meeting as an observer, and have begun (phone chat last week) discussion with Tom Baker (chair) about potential for having a common model across both FOAF and DC. Tom is also leading the SW Best Practices taskforce on Vocabulary Management at W3C, so there's potential there I think. I'll try to find out what DC use, re last-modified etc. > "term documentation" > I volunteered at foafcamp to build some kind of a > "term browser" for easier access to the description > of selected foaf terms. Each term view could > show a term's standard information (label, comment, > term_status, relations, description, rdf/xml), but > also more detailed stuff such as examples, links to > tools, use cases, best practices, similar/related > terms, version info, issues, todos, etc. > Any thoughts on useful documentation types? Sounds cool! Translations too? :) > Just some ideas.. Thanks. See you later... Dan From Libby.Miller at bristol.ac.uk Wed Sep 22 18:44:54 2004 From: Libby.Miller at bristol.ac.uk (Libby Miller) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] foaf:dateOfBirth In-Reply-To: <20040921224230.GP8052@soto.kasei.com> References: <479B5AE2-E3C7-11D8-9F63-000D932A33F0@atommic.com> <20040921224230.GP8052@soto.kasei.com> Message-ID: hi Tony, hm, interesting point. At the moment you would have to talk about them separately, e.g. or somesuch. I'm not saying this is the right way mind you :) Plus it doesn't really say what you want....but it sort of provides a kind of evidence that we met, much like foaf:depicts does. RDF ical is at http://www.w3.org/2002/12/cal/ (currently slightly in flux) Libby On Tue, 21 Sep 2004, Tony Bowden wrote: > On Sun, Aug 01, 2004 at 03:46:31PM +0100, Libby Miller wrote: >> You could also use the bio vocab: >> [[ >> >> >> 1970-06-15 >> Brentwood, Essex, United Kingdom >> >> >> ]] >> http://vocab.org/bio/0.1/#Birth > > Are there other vocabs that include a wider range of events? Rather than > just saying that I foaf:knows someone, I'd often prefer to say that I > met them at a certain event. What's the preferred way to do something > like this? > > Tony > > > _______________________________________________ > rdfweb-dev mailing list > rdfweb-dev@vapours.rdfweb.org > wiki: http://rdfweb.org/topic/FoafProject > http://rdfweb.org/mailman/listinfo/rdfweb-dev > From bnowack at appmosphere.com Wed Sep 22 19:47:35 2004 From: bnowack at appmosphere.com (Benjamin Nowack) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] proposals for enhancing the descriptions of foaf terms Message-ID: [[ 17:04:57 action: bengee propose some clarifications re lifecycle/stability vocab to list, goal of having better machine-readable status for FOAF namespace ]] (from 2004-09-22's IRC chat [1]) There are actually two proposals we could discuss on this list (whether they make sense at all, how/if they could be implemented, etc.): 1) The foaf spec should give more info about a term's lifecycle stage than it is currently done via the "unstable"/"testing"/ "stable" term_status annotations. At the moment, it's not possible to see, e.g. *when* a term's status went to "testing". Or how long a term has been "unstable" (which could maybe tell a tool developer how actively it is maintained/how likely it is to move to "stable", etc). 2) the FOAF namespace should distinguish terms which "may be removed from spec", from terms whose "usage is discouraged" (because there are better idioms to use), but which will probably stay in the namespace indefinitely. some thoughts: re 1) We could use prose w/ owl:versionInfo, but having some kind of machine-readable "lastmodified" annotation would e.g. allow auto-generating an RSS feed for updated/added terms. This would also be possible if we used an agreed-on date/time format and owl:versionInfo. perhaps a DC term could be used. re 2) owl:Deprecated[Class|Property] could cover at least one of the cases. The owl reference doc says "by deprecating a term, it means that the term should not be used in new documents that commit to the ontology". As "may be removed from the spec" somehow includes "usage is discouraged", we possibly don't even need to distinguish the cases. A mentioned alternative would be the use of recommended term subsets (aka profiles ;) for different use cases or application areas. A re-worded proposal could then be "the foaf terms should have (machine-readable) pointers to application areas/use cases/implementing apps" which could allow the automatic generation of subsets for given use cases, or term sets of widely deployed terms." ideas, comments, objections? /action item bengee [1] http://www.ilrt.bris.ac.uk/discovery/chatlogs/rdfig/2004-09-22.html#T17-04-57 benjamin -- Benjamin Nowack Kruppstr. 100 45145 Essen, Germany http://www.appmosphere.com/ From bnowack at appmosphere.com Wed Sep 22 19:59:14 2004 From: bnowack at appmosphere.com (Benjamin Nowack) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] agenda item suggestion for wednesday's chat In-Reply-To: <20040922142359.GE5627@homer.w3.org> References: <20040922142359.GE5627@homer.w3.org> Message-ID: On 22.09.2004 10:23:59, Dan Brickley wrote: >"* Benjamin Nowack" <...> >> I volunteered at foafcamp to build some kind of a >> "term browser" for easier access to the description >> of selected foaf terms. Each term view could >> show a term's standard information (label, comment, >> term_status, relations, description, rdf/xml), but >> also more detailed stuff such as examples, links to >> tools, use cases, best practices, similar/related >> terms, version info, issues, todos, etc. >> Any thoughts on useful documentation types? > >... >Translations too? :) the tool supports different languages for class/property/ vocabulary annotations, and class/property doc items, but I'm not sure about character set limitations. At least chars from iso-8859-1 should work. benjamin >Thanks. See you later... >Dan > -- Benjamin Nowack Kruppstr. 100 45145 Essen, Germany http://www.appmosphere.com/ From tony-rdf at kasei.com Wed Sep 22 21:04:35 2004 From: tony-rdf at kasei.com (Tony Bowden) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] foaf:dateOfBirth In-Reply-To: References: <479B5AE2-E3C7-11D8-9F63-000D932A33F0@atommic.com> <20040921224230.GP8052@soto.kasei.com> Message-ID: <20040922210434.GA6109@soto.kasei.com> On Wed, Sep 22, 2004 at 07:44:54PM +0100, Libby Miller wrote: > hm, interesting point. At the moment you would have to talk about them > separately > > > > > > or somesuch. I'm not saying this is the right way mind you :) > Plus it doesn't really say what you want....but it sort of provides a > kind of evidence that we met, much like foaf:depicts does. I doesn't really provide much evidence though ... FOAFcamp was pretty small and I'm sure there were people there that I didn't meet. Start widening it out to something like OsCon, or even a sporting event, or concert or the like, and the chances that 2 people who were attendees actually met start getting much slimmer ... Co-depictions are slightly better, as there's a decent chance that 2 people in the same photograph (assuming it's of a small enough number of people) have indeed met [although I do have a bizarre story about how I once discovered than an ex-girlfriend had a photograph of the two of us from several years before we'd actually met ...] I'll certainly explore the ical stuff a bit more, though. Thanks, Tony Tony From mof-rdf at mfd-consult.dk Thu Sep 23 22:06:19 2004 From: mof-rdf at mfd-consult.dk (Morten Frederiksen) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] domain of foaf:interest Message-ID: <200409240006.19336.mof-rdf@mfd-consult.dk> Hi, Perhaps an agenda item for the next ScheduledTopicChat: When hacking up an "author list" for the WP FOAF plugin [1], I noticed that the blogroll of planetrdf [2] had some foaf:interest properties on the foaf:Document nodes describing the weblogs that are aggregated. This leads to a textbook contradiction [3], as the domain of foaf:interest is foaf:Person, and foaf:Document and foaf:Person are declared as owl:disjointWith each other. After chatting a bit with dajobe, who changed the blogroll right away, we realised that the change wasn't enough, as the blogroll contains weblogs of groups and projects, some of which may be said to have interests. However, because of the domain "contraint", these are now all person's, which is clearly wrong. One solution would be to remove the interest statements, perhaps using foaf:topic on the channels/items, but it doesn't seem to be all that irrelevant to talk of interests for at least groups, so I suggest relaxing the domain of foaf:interest to foaf:Agent instead of foaf:Person. [1] http://www.wasab.dk/morten/blog/archives/2004/07/05/wordpress-plugin-foaf-output [2] http://journal.dajobe.org/journal/2003/07/semblogs/bloggers.rdf [3] http://rdfweb.org/topic/FoafContradictions Regards, Morten From danbri at w3.org Fri Sep 24 00:19:12 2004 From: danbri at w3.org (Dan Brickley) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] domain of foaf:interest In-Reply-To: <200409240006.19336.mof-rdf@mfd-consult.dk> References: <200409240006.19336.mof-rdf@mfd-consult.dk> Message-ID: <20040924001912.GF1975@homer.w3.org> * Morten Frederiksen [2004-09-24 00:06+0200] > Hi, > > Perhaps an agenda item for the next ScheduledTopicChat: > > When hacking up an "author list" for the WP FOAF plugin [1], I noticed that > the blogroll of planetrdf [2] had some foaf:interest properties on the > foaf:Document nodes describing the weblogs that are aggregated. > > This leads to a textbook contradiction [3], as the domain of foaf:interest is > foaf:Person, and foaf:Document and foaf:Person are declared as > owl:disjointWith each other. > > After chatting a bit with dajobe, who changed the blogroll right away, we > realised that the change wasn't enough, as the blogroll contains weblogs of > groups and projects, some of which may be said to have interests. However, > because of the domain "contraint", these are now all person's, which is > clearly wrong. > > One solution would be to remove the interest statements, perhaps using > foaf:topic on the channels/items, but it doesn't seem to be all that > irrelevant to talk of interests for at least groups, so I suggest relaxing > the domain of foaf:interest to foaf:Agent instead of foaf:Person. Good point. And who says OWL has only esoteric uses! It helps keep our modeling honest, at least... Anyone here object to such a change? If I don't here complaints, it'll happen. Dan > > > [1] > http://www.wasab.dk/morten/blog/archives/2004/07/05/wordpress-plugin-foaf-output > [2] http://journal.dajobe.org/journal/2003/07/semblogs/bloggers.rdf > [3] http://rdfweb.org/topic/FoafContradictions > > > Regards, > Morten > > _______________________________________________ > rdfweb-dev mailing list > rdfweb-dev@vapours.rdfweb.org > wiki: http://rdfweb.org/topic/FoafProject > http://rdfweb.org/mailman/listinfo/rdfweb-dev From james.carlyle at takepart.com Fri Sep 24 14:12:16 2004 From: james.carlyle at takepart.com (James Carlyle) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] domain of foaf:interest In-Reply-To: <41542914.5050804@semanticplanet.com> References: <41542914.5050804@semanticplanet.com> Message-ID: <41542B40.5090207@takepart.com> Hi I would also be interested in discussing foaf:interest at the next ScheduledTopicChat. Ian Davis and I are building FriendSpace, trying to foster the concept "Community of Common Interests". We are bootstrapping off existing FOAF usage and the idea of "user-owned data", and are using foaf:interest to express voluntary membership of the group. We have 2 issues: 1) We want to present an aggregation of the following kind of personal activity for the community: Weblog posts, calendar events, bookmarks and possibly relationships (foaf:knows) Current foaf:interest usage describes the interests of a Person on a broad scale, but not the facets of a person's activities. So if we aggregate the daily activities of members of the group, we have no way of only looking for weblog posts or events that relate to a particular interest. In other words we need a way of finding how individual weblog posts, calendar events, bookmarks and foaf:knows relate to the foaf:interest that they say they have. There seem to be 2 alternatives here - see if people are willing to publish an interest-focused weblog, for example (or a category-focused sub-blog) i.e. Person interest Document (say uri is http://rdfweb.org/) made Channel page (domain is Resource, range is Document) Document (say uri is http://rdfweb.org/) made Channel (a.n.other) page ...a.n.other resource or to tag individual items with a "topic" i.e. Person interest Document (say uri is http://rdfweb.org/) seeAlso Vcalendar component Vevent page Document (say uri is http://rdfweb.org/) 2) The range of foaf:interest is foaf:Document, but it seems intuitive to say that I am interested in both tangible things, and intangible things that might not be reflected easily by "indicating a foaf:Document whose foaf:topic(s) broadly characterises that interest" (quoting from the FOAF spec). I am sure there is a lot of historical context here that I can't find, and would be interested in hearing the reasoning for this. I know Leigh, Morten and Kanzaki-san had views on the mailing list [1]. As you can see, we could tie a Vevent to a Document via foaf:page, but this seems an odd way of characterising RSS items or Vevents. Is there a better way? Kind regards James Carlyle [1] http://rdfweb.org/pipermail/rdfweb-dev/2004-March/012871.html >* Morten Frederiksen [2004-09-24 00:06+0200] > >>Hi, >> >>Perhaps an agenda item for the next ScheduledTopicChat: >> >>When hacking up an "author list" for the WP FOAF plugin [1], I noticed that >>the blogroll of planetrdf [2] had some foaf:interest properties on the >>foaf:Document nodes describing the weblogs that are aggregated. >> From mof-rdf at mfd-consult.dk Fri Sep 24 15:16:04 2004 From: mof-rdf at mfd-consult.dk (Morten Frederiksen) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] domain of foaf:interest In-Reply-To: <41542914.5050804@semanticplanet.com> References: <41542914.5050804@semanticplanet.com> Message-ID: <200409241716.04096.mof-rdf@mfd-consult.dk> Hey, On Friday 24 September 2004 16:03, James Carlyle wrote: > 1) We want to present an aggregation of the following kind of personal > activity for the community: > Weblog posts, calendar events, bookmarks and possibly relationships > (foaf:knows) This is great, just the kind of tools we need. > There seem to be 2 alternatives here - see if people are willing to > publish an interest-focused weblog, for example (or a category-focused > sub-blog) I tried to do this with my FOAF output plugin for WP [1] (here I go, plugging it again), which also "enhances" the RSS feed [2] with topic information, like this: rss:item topic Document (X) At the same time, the FOAF file contains statements of interest: Person interest Document (X) The topic/interest information is based on categories and URIs assigned to them. > 2) The range of foaf:interest is foaf:Document, but it seems intuitive > to say that I am interested in both tangible things, and intangible > things that might not be reflected easily by "indicating a foaf:Document > whose foaf:topic(s) broadly characterises that interest" (quoting from > the FOAF spec). I am sure there is a lot of historical context here > that I can't find, and would be interested in hearing the reasoning for > this. I won't claim to understand the entire history, but one point is that foaf:topic and foaf:interest fit nicely together for exactly reasons like this, to be able to create connections. Remember that foaf:interest doesn't entail interest in the document, but in the *topic* of the document. Regards, Morten From danny.ayers at gmail.com Sat Sep 25 18:22:51 2004 From: danny.ayers at gmail.com (Danny Ayers) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] foaf:dateOfBirth In-Reply-To: <20040921224230.GP8052@soto.kasei.com> References: <479B5AE2-E3C7-11D8-9F63-000D932A33F0@atommic.com> <20040921224230.GP8052@soto.kasei.com> Message-ID: <1f2ed5cd04092511223b30fcc5@mail.gmail.com> On Tue, 21 Sep 2004 23:42:30 +0100, Tony Bowden wrote: > Are there other vocabs that include a wider range of events? Rather than > just saying that I foaf:knows someone, I'd often prefer to say that I > met them at a certain event. What's the preferred way to do something > like this? I like the sound of what this suggests - xxx:met may be *too* simple, though it could be used with subclassing like foaf:knows. But given the existing vocabs for space/time etc I think it would probably be better /not/ to qualify it with where and when, etc. rather than having xxx:metAtBusStop What about a class something like Encounter, maybe (not quite the right properties, but you know what I mean): Encounter geo:location #place ical:when #date ical:occasion #event xxx:participant foaf:Person xxx:participant foaf:Person "participant" seems a bit clunky but I haven't the energy to click for a thesaurus. Cheers, Danny. From tony-rdf at kasei.com Sat Sep 25 19:18:20 2004 From: tony-rdf at kasei.com (Tony Bowden) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] foaf:dateOfBirth In-Reply-To: <1f2ed5cd04092511223b30fcc5@mail.gmail.com> References: <479B5AE2-E3C7-11D8-9F63-000D932A33F0@atommic.com> <20040921224230.GP8052@soto.kasei.com> <1f2ed5cd04092511223b30fcc5@mail.gmail.com> Message-ID: <20040925191820.GA10992@soto.kasei.com> On Sat, Sep 25, 2004 at 08:22:51PM +0200, Danny Ayers wrote: > But given the existing vocabs for space/time etc I > think it would probably be better /not/ to qualify it with where and > when, etc. rather than having xxx:metAtBusStop I'm not sure I follow what this means ... > What about a class something like Encounter, maybe (not quite the > right properties, but you know what I mean): > Encounter > geo:location > #place > ical:when > #date > ical:occasion > #event > xxx:participant > foaf:Person > xxx:participant > foaf:Person At first glance looks good! Tony From james.carlyle at takepart.com Mon Sep 27 12:22:58 2004 From: james.carlyle at takepart.com (James Carlyle) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] domain of foaf:interest In-Reply-To: <200409241716.04096.mof-rdf@mfd-consult.dk> References: <41542914.5050804@semanticplanet.com> <200409241716.04096.mof-rdf@mfd-consult.dk> Message-ID: <41580622.4080302@takepart.com> Morten Thanks for your encouragement and for your work on the WP Foaf plugin. Just a small point - shouldn't rss:item have a property called foaf:page instead of foaf:topic which points to a foaf:Document? From the Foaf spec, foaf:Document is in the range of foaf:interest and foaf:page, and the domain (not the range) of foaf:topic is foaf:Document. I am assuming you want to indicate how the interest of the foaf:Person relates to the rss:item. James > >I tried to do this with my FOAF output plugin for WP [1] (here I go, plugging >it again), which also "enhances" the RSS feed [2] with topic information, >like this: > >rss:item > topic > Document (X) > >At the same time, the FOAF file contains statements of interest: > >Person > interest > Document (X) > >The topic/interest information is based on categories and URIs assigned to >them. > > From iand at internetalchemy.org Mon Sep 27 13:26:45 2004 From: iand at internetalchemy.org (Ian Davis) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] foaf:interest et al. Message-ID: <41581515.4060601@internetalchemy.org> Hi all, I've been having a think about foaf:interest and its close buddies foaf:topic and foaf:interest_topic which I've written about here[1] I think the foaf:interest property is misnamed. The problem is that the name is very general but the meaning is very specific (a page about a topic of interest) and somewhat at a tangent to the common meaning of the word. Not only is this confusing for authors it makes it harder to introduce new properties that are more general than foaf:interest. For example, I think the intended meaning of foaf:interest_topic (A thing of interest to this person.) is closer to the common meaning of interest. There's a lot of deployed FOAF out there that uses foaf:interest so this makes it very hard to change. However, it would be good to rationalise this part of the spec with a view to deprecating foaf:interest one day. My suggestion is to rename the foaf:topic_interest property to foaf:topicOfInterest or foaf:interestInTopic with the definition: A topic that this person expresses an interest in. Then, for completeness I'd also suggest the introduction of a foaf:Topic class. The domain of foaf:topicOfInterest and foaf:topic would be foaf:Topic. The range of foaf:topicOfInterest would be foaf:Agent. A better name for foaf:interest is harder, in my posting I suggest rather tongue in cheek foaf:interestInWhateverThisPageIsAbout. Maybe pageAboutInterest or interestPage or, to parallel a suggestion above, interestInPageTopic (i.e. bob has interest in this page's topic). Ian [1] http://www.semanticplanet.com/2004/09/topics.html From john.breslin at deri.org Mon Sep 27 14:28:37 2004 From: john.breslin at deri.org (John Breslin) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] foaf:interest et al. In-Reply-To: <41581515.4060601@internetalchemy.org> Message-ID: <415831A5.16307.1820A02@localhost> If the foaf:topicOfInterest pointed to a category in a structure like the ODP, then maybe the problem of requiring a foaf:interestInWhateverThisPageIsAbout could be done away with :) I remember someone in FOAF Galway using foaf:interest with both a label and a URL - this seemed like good practice. I also wanted to use foaf:interest or foaf:topic_interest as part of my bulletin board FOAF export; people can define a like or dislike as shown here: http://www.boards.jp/network/likedis.php?do=view&u=1 where a label and URL to an ODP category are required, and another non-ODP URL is optional. But this leads me to ask how it would be possible to define a negative interest in something. This could be handy if for example I said that I had an interest in Sports, but do not have an interest in Baseball - so if things are recommended to me in a Sports domain, leave out the Baseball ones. foaf:topicOfNoInterest anyone? John. -- Dr. John Breslin Digital Enterprise Research Institute http://www.johnbreslin.com/ john.breslin@deri.org From iand at internetalchemy.org Mon Sep 27 14:50:18 2004 From: iand at internetalchemy.org (Ian Davis) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] foaf:interest et al. In-Reply-To: <415831A5.16307.1820A02@localhost> References: <415831A5.16307.1820A02@localhost> Message-ID: <415828AA.8020604@internetalchemy.org> On 27/09/2004 15:28, John Breslin wrote: > > I remember someone in FOAF Galway using foaf:interest with both a label and a > URL - this seemed like good practice. This is good practice, but the label would apply to the page not to the topic. You could assign a label to the topic like this: The topic label > > But this leads me to ask how it would be possible to define a negative interest > in something. This could be handy if for example I said that I had an interest > in Sports, but do not have an interest in Baseball - so if things are > recommended to me in a Sports domain, leave out the Baseball ones. > foaf:topicOfNoInterest anyone? It should be possible to use Owl to do this by defining a class that is the the intersection of Sport with the complement of Baseball. Defining those two classes is left as an exercise for the reader... Ian From ndw at nwalsh.com Mon Sep 27 14:58:19 2004 From: ndw at nwalsh.com (Norman Walsh) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] FOAF infrastructure changes requested Message-ID: <87vfdzkbck.fsf@nwalsh.com> Hello world, While working to improve the metadata presentation on norman.walsh.name, I stumbled over two issues. 1. The FOAF RDF Schema and the FOAF spec are content-negotiated from the namespace URI (http://xmlns.com/foaf/0.1/). This means that a tool like XSLT's document() function cannot get the schema from the namespace URI. I humbly suggest that the namespace URI should return the schema with a stylesheet PI that does something reasonable and includes a pointer to the human-readable spec. 2. The spec begins: ... I humbly request that it use one-fewer RDF serialization trick :-) and instead be coded as: Friend of a Friend (FOAF) vocabulary The Friend of a Friend (FOAF) RDF vocabulary, described using W3C RDF Schema and the Web Ontology Language. $Date: 2004/09/01 15:37:56 $ ... Granted, I'm asking because it makes my life easier, but I figure it can't hurt to ask :-) Be seeing you, norm -- Norman Walsh | Space isn't remote at all. It's only an http://nwalsh.com/ | hour's drive away if your car could go | straight upwards.--Fred Hoyle -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 188 bytes Desc: not available Url : http://lists.usefulinc.com/pipermail/foaf-dev/attachments/20040927/f7ebdc3b/attachment-0001.pgp From iand at internetalchemy.org Mon Sep 27 15:09:03 2004 From: iand at internetalchemy.org (Ian Davis) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] FOAF infrastructure changes requested In-Reply-To: <87vfdzkbck.fsf@nwalsh.com> References: <87vfdzkbck.fsf@nwalsh.com> Message-ID: <41582D0F.7020203@internetalchemy.org> Hi Norman, On 27/09/2004 15:58, Norman Walsh wrote: > While working to improve the metadata presentation on norman.walsh.name, > I stumbled over two issues. > > 1. The FOAF RDF Schema and the FOAF spec are content-negotiated from > the namespace URI (http://xmlns.com/foaf/0.1/). This means that a tool > like XSLT's document() function cannot get the schema from the > namespace URI. > > I humbly suggest that the namespace URI should return the schema with > a stylesheet PI that does something reasonable and includes a pointer > to the human-readable spec. Can I suggest an alternative? If you're dispatching based on namespace, perhaps you could maintain a simple namespace/schema mapping file, e.g.: You might be able to do something like: document( document( mappings.rdf )/rdf:Description[ @rdf:about="..."]/rdfs:seeAlso[0]/@rdf:resource ) Ian From jim.ley at gmail.com Mon Sep 27 15:18:36 2004 From: jim.ley at gmail.com (Jim Ley) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] FOAF infrastructure changes requested In-Reply-To: <87vfdzkbck.fsf@nwalsh.com> References: <87vfdzkbck.fsf@nwalsh.com> Message-ID: <851c8d3104092708186fb460ea@mail.gmail.com> On Mon, 27 Sep 2004 10:58:19 -0400, Norman Walsh wrote: > 1. The FOAF RDF Schema and the FOAF spec are content-negotiated from > the namespace URI (http://xmlns.com/foaf/0.1/). This means that a tool > like XSLT's document() function cannot get the schema from the > namespace URI. I don't quite see why this follows, surely an XSLT parser must send an accept header containing only XML mime-types as acceptable. whilst including application/rdf+xml may be too much to expect, ensuring that the foaf URI returns XML in response to application/xml aswell should be fine. > I humbly suggest that the namespace URI should return the schema with > a stylesheet PI that does something reasonable and includes a pointer > to the human-readable spec. I really would rather not have XSLT PI's in the foaf spec - to make my life easier of course - I also prefer the human readable HTML as the default format. Have the best practices people looked at this sort of stuff yet? Cheers, Jim. From ndw at nwalsh.com Mon Sep 27 15:26:27 2004 From: ndw at nwalsh.com (Norman Walsh) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] Re: FOAF infrastructure changes requested In-Reply-To: <41582D0F.7020203@internetalchemy.org> References: <87vfdzkbck.fsf@nwalsh.com> <41582D0F.7020203@internetalchemy.org> Message-ID: <877jqfhgws.fsf@nwalsh.com> LyBJYW4gRGF2aXMgPGlhbmRAaW50ZXJuZXRhbGNoZW15Lm9yZz4gd2FzIGhlYXJkIHRvIHNheToK fCBIaSBOb3JtYW4sCnwKfCBPbiAyNy8wOS8yMDA0IDE1OjU4LCBOb3JtYW4gV2Fsc2ggd3JvdGU6 Cnw+IFdoaWxlIHdvcmtpbmcgdG8gaW1wcm92ZSB0aGUgbWV0YWRhdGEgcHJlc2VudGF0aW9uIG9u IG5vcm1hbi53YWxzaC5uYW1lLAp8PiBJIHN0dW1ibGVkIG92ZXIgdHdvIGlzc3Vlcy4KfD4gMS4g VGhlIEZPQUYgUkRGIFNjaGVtYSBhbmQgdGhlIEZPQUYgc3BlYyBhcmUgY29udGVudC1uZWdvdGlh dGVkIGZyb20KfD4gICAgdGhlIG5hbWVzcGFjZSBVUkkgKGh0dHA6Ly94bWxucy5jb20vZm9hZi8w LjEvKS4gVGhpcyBtZWFucyB0aGF0IGEgdG9vbAp8PiAgICBsaWtlIFhTTFQncyBkb2N1bWVudCgp IGZ1bmN0aW9uIGNhbm5vdCBnZXQgdGhlIHNjaGVtYSBmcm9tIHRoZQp8PiAgICBuYW1lc3BhY2Ug VVJJLgp8PiAgICBJIGh1bWJseSBzdWdnZXN0IHRoYXQgdGhlIG5hbWVzcGFjZSBVUkkgc2hvdWxk IHJldHVybiB0aGUgc2NoZW1hCnw+IHdpdGgKfD4gICAgYSBzdHlsZXNoZWV0IFBJIHRoYXQgZG9l cyBzb21ldGhpbmcgcmVhc29uYWJsZSBhbmQgaW5jbHVkZXMgYSBwb2ludGVyCnw+ICAgIHRvIHRo ZSBodW1hbi1yZWFkYWJsZSBzcGVjLgp8CnwgQ2FuIEkgc3VnZ2VzdCBhbiBhbHRlcm5hdGl2ZT8g SWYgeW91J3JlIGRpc3BhdGNoaW5nIGJhc2VkIG9uCnwgbmFtZXNwYWNlLCBwZXJoYXBzIHlvdSBj b3VsZCBtYWludGFpbiBhIHNpbXBsZSBuYW1lc3BhY2Uvc2NoZW1hCnwgbWFwcGluZyBmaWxlLCBl LmcuOgp8CnwgPHJkZjpEZXNjcmlwdGlvbiByZGY6YWJvdXQ9Imh0dHA6Ly94bWxucy5jb20vZm9h Zi8wLjEvIj4KfCAgICA8cmRmczpzZWVBbHNvIHJkZjpyZXNvdXJjZT0iaHR0cDovL3htbG5zLmNv bS9mb2FmLzAuMS9pbmRleC5yZGYiLz4KfCA8L3JkZjpEZXNjcmlwdGlvbj4KfAp8IFlvdSBtaWdo dCBiZSBhYmxlIHRvIGRvIHNvbWV0aGluZyBsaWtlOgp8CnwgZG9jdW1lbnQoIGRvY3VtZW50KCBt YXBwaW5ncy5yZGYgKS9yZGY6RGVzY3JpcHRpb25bCnwgQHJkZjphYm91dD0iLi4uIl0vcmRmczpz ZWVBbHNvWzBdL0ByZGY6cmVzb3VyY2UgKQoKSSBjb3VsZCBkbyB0aGF0LCBidXQgSSBzdGlsbCBw cmVmZXIgbXkgYWx0ZXJuYXRpdmUuCgogICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAg ICAgICAgQmUgc2VlaW5nIHlvdSwKICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAg ICAgICAgbm9ybQoKLS0gCk5vcm1hbiBXYWxzaCA8bmR3QG53YWxzaC5jb20+IHwgQWxsIG91ciBm b2VzIGFyZSBtb3J0YWwuLS1QYXVsIFZhbMOpcnkKaHR0cDovL253YWxzaC5jb20vICAgICAgICAg ICAgfCAKLS0tLS0tLS0tLS0tLS0gbmV4dCBwYXJ0IC0tLS0tLS0tLS0tLS0tCkEgbm9uLXRleHQg YXR0YWNobWVudCB3YXMgc2NydWJiZWQuLi4KTmFtZTogbm90IGF2YWlsYWJsZQpUeXBlOiBhcHBs aWNhdGlvbi9wZ3Atc2lnbmF0dXJlClNpemU6IDE4OCBieXRlcwpEZXNjOiBub3QgYXZhaWxhYmxl ClVybCA6IGh0dHA6Ly9saXN0cy51c2VmdWxpbmMuY29tL3BpcGVybWFpbC9mb2FmLWRldi9hdHRh Y2htZW50cy8yMDA0MDkyNy9lMGEzMTE2Ny9hdHRhY2htZW50LTAwMDEucGdwCg== From mof-rdf at mfd-consult.dk Mon Sep 27 15:58:20 2004 From: mof-rdf at mfd-consult.dk (Morten Frederiksen) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] domain of foaf:interest In-Reply-To: <41580622.4080302@takepart.com> References: <41542914.5050804@semanticplanet.com> <200409241716.04096.mof-rdf@mfd-consult.dk> <41580622.4080302@takepart.com> Message-ID: <200409271758.20428.mof-rdf@mfd-consult.dk> Hi James, all, On Monday 27 September 2004 14:22, James Carlyle wrote: > Just a small point - shouldn't rss:item have a property called foaf:page > instead of foaf:topic which points to a foaf:Document? From the Foaf > spec, foaf:Document is in the range of foaf:interest and foaf:page, and > the domain (not the range) of foaf:topic is foaf:Document. Right you are, wrong I was... Fortunately, I misrepresented what I had actually done (well, I had another error in there, but that was pointed out to me by Ian today [1]), so what I actually have is this [2]: rss:item topic (Something) dc:title page Document (X) Which, even if not exactly what you propose, should be OK (see also Ian's excellent writeup on the matter), except perhaps for the dc:title property, which perhaps should be rdfs:label or even foaf:name, as use of the Dublin Core properties seems best suited for documents. Opinions welcome. Sorry for the confusion. [1] http://www.wasab.dk/morten/blog/archives/2004/05/20/improving-rss-output-from-wordpress#comment-179 [2] http://www.wasab.dk/morten/blog/feed/rdf Regards, Morten From gk at ninebynine.org Mon Sep 27 14:28:43 2004 From: gk at ninebynine.org (Graham Klyne) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] "Agent Model Yields Leadership" Message-ID: <5.1.0.14.2.20040927152553.02f3c948@127.0.0.1> This turned up in the ACM Technews digest: http://www.acm.org/technews/articles/2004-6/0922w.html#item5 The referenced article seems to suggest some interesting possibilities for FOAF-like networks. There are also resonances here with some of the trust-related work I've read about. #g -- # "Agent Model Yields Leadership" Technology Research News (09/29/04); Patch, Kimberly Researchers at Los Alamos National Laboratory and two universities have developed a software model for studying economic markets, quantitative sociology, or optimizing communications among robot collectives. The model is based on the classic minority game, where multiple agents compete to be in the minority of each round of decision-making; by adding a limited social network between the agents, the researchers were able to create a leadership structure that ultimately led to smarter and more adaptive performance than classic models without social network influence. Large and complicated systems such as the stock market are difficult to model because of the number of independent agents and choices available. Computing all possible scenarios is impossible with today's technology, but the researchers' model uses quantitative representations for agent behavior. Los Alamos National Laboratory researcher Zoltan Toroczkai says real human agents actually make decisions inductively rather than through deductive reasoning, as is assumed in classic game theory models; this is because, as with the computer models, figuring out all the possibilities is simply too difficult. The social network influence links each agent to its nearest neighbors and has them rely on the most recently successful agents for advice. Interestingly, the model grew more volatile when denser connectivity was added, since some leader agents' opinions became too popular and destabilized the system. Eventually, Toroczkai says the software model could help arrays of robots operate in conjunction where no human control is possible, such as on Mars explorations, although this type of technology would not be ready for another 10 to 20 years. The research was funded by the National Science Foundation, the Department of Energy, the Research Corporation, and the Alfred P. Sloan Foundation. Click Here to View Full Article http://www.trnmag.com/Stories/2004/092204/Agent_model_yields_leadership_092204.html ------------ Graham Klyne For email: http://www.ninebynine.org/#Contact From ndw at nwalsh.com Mon Sep 27 17:35:02 2004 From: ndw at nwalsh.com (Norman Walsh) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] Re: FOAF infrastructure changes requested In-Reply-To: <851c8d3104092708186fb460ea@mail.gmail.com> References: <87vfdzkbck.fsf@nwalsh.com> <851c8d3104092708186fb460ea@mail.gmail.com> Message-ID: <87is9z8vjt.fsf@nwalsh.com> / Jim Ley was heard to say: | On Mon, 27 Sep 2004 10:58:19 -0400, Norman Walsh wrote: |> 1. The FOAF RDF Schema and the FOAF spec are content-negotiated from |> the namespace URI (http://xmlns.com/foaf/0.1/). This means that a tool |> like XSLT's document() function cannot get the schema from the |> namespace URI. | | I don't quite see why this follows, surely an XSLT parser must send an | accept header containing only XML mime-types as acceptable. whilst | including application/rdf+xml may be too much to expect, ensuring that | the foaf URI returns XML in response to application/xml aswell should | be fine. Alas, XSLT provides no way to influence the accept headers sent by the implementation. I imagine that application/xml is sent, but the server considers the XHTML version a better match. |> I humbly suggest that the namespace URI should return the schema with |> a stylesheet PI that does something reasonable and includes a pointer |> to the human-readable spec. | | I really would rather not have XSLT PI's in the foaf spec - to make my | life easier of course - I also prefer the human readable HTML as the | default format. Have the best practices people looked at this sort of | stuff yet? Dunno. I'm not sure content negotiating RDF and human text is a good idea, but I seem to be in the minority. Be seeing you, norm -- Norman Walsh | The Future is something which everyone http://nwalsh.com/ | reaches at the rate of sixty minutes an | hour, whatever he does, whoever he | is.--C. S. Lewis -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 188 bytes Desc: not available Url : http://lists.usefulinc.com/pipermail/foaf-dev/attachments/20040927/d60d2483/attachment-0001.pgp From iand at internetalchemy.org Mon Sep 27 18:14:45 2004 From: iand at internetalchemy.org (Ian Davis) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] FOAF infrastructure changes requested In-Reply-To: <851c8d3104092708186fb460ea@mail.gmail.com> References: <87vfdzkbck.fsf@nwalsh.com> <851c8d3104092708186fb460ea@mail.gmail.com> Message-ID: <41585895.40505@internetalchemy.org> On 27/09/2004 16:18, Jim Ley wrote: > On Mon, 27 Sep 2004 10:58:19 -0400, Norman Walsh wrote: > >>1. The FOAF RDF Schema and the FOAF spec are content-negotiated from >> the namespace URI (http://xmlns.com/foaf/0.1/). This means that a tool >> like XSLT's document() function cannot get the schema from the >> namespace URI. > > > I don't quite see why this follows, surely an XSLT parser must send an > accept header containing only XML mime-types as acceptable. whilst > including application/rdf+xml may be too much to expect, ensuring that > the foaf URI returns XML in response to application/xml aswell should > be fine. This might be problematic since application/xhtml+xml could also be returned. Both are obviously acceptable to an XSLT parser but they have differing levels of usefulness :) Ian From danbri at w3.org Mon Sep 27 18:56:53 2004 From: danbri at w3.org (Dan Brickley) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] FOAF infrastructure changes requested In-Reply-To: <87vfdzkbck.fsf@nwalsh.com> References: <87vfdzkbck.fsf@nwalsh.com> Message-ID: <20040927185653.GA15164@homer.w3.org> * Norman Walsh [2004-09-27 10:58-0400] > Hello world, > > While working to improve the metadata presentation on norman.walsh.name, > I stumbled over two issues. > > 1. The FOAF RDF Schema and the FOAF spec are content-negotiated from > the namespace URI (http://xmlns.com/foaf/0.1/). This means that a tool > like XSLT's document() function cannot get the schema from the > namespace URI. > > I humbly suggest that the namespace URI should return the schema with > a stylesheet PI that does something reasonable and includes a pointer > to the human-readable spec. Do you think XSLT is deployable for such things yet? CSS styling can't, AFAIK, generate hyperlinks, so would be a hypertextual dead-end. If/when XSLT is "out there" enough, and browsers default to running untrusted XSLTs, then it could be the way to go. > 2. The spec begins: > > dc:title="Friend of a Friend (FOAF) vocabulary" > dc:description="The Friend of a Friend (FOAF) RDF vocabulary, described using W3C RDF Schema and the Web Ontology Language." > dc:date="$Date: 2004/09/01 15:37:56 $"> > ... > > I humbly request that it use one-fewer RDF serialization trick :-) and The reason we have an attribute-centric serialization is so the RDF can be shoved inside XHTML and the content not spill out in downlevel browsers. For the pure RDF version we could publish an index.rdf which had an element-based encoding. If someone wants to write the XSLT that generates one from the other, I could build it into the check scripts. > instead be coded as: > > > Friend of a Friend (FOAF) vocabulary > The Friend of a Friend (FOAF) RDF vocabulary, described using W3C RDF Schema and the Web Ontology Language. > $Date: 2004/09/01 15:37:56 $ > ... > > Granted, I'm asking because it makes my life easier, but I figure it can't > hurt to ask :-) FOAF is very much an environment where we can experiment with figuring out the options for what we rig up for de-referencable namespaces. I'm not sure there are many obvious right answers yet. Am happy to evolve things to try to meet as many people's needs as possible... cheers, Dan > _______________________________________________ > rdfweb-dev mailing list > rdfweb-dev@vapours.rdfweb.org > wiki: http://rdfweb.org/topic/FoafProject > http://rdfweb.org/mailman/listinfo/rdfweb-dev From danbri at w3.org Mon Sep 27 19:00:09 2004 From: danbri at w3.org (Dan Brickley) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] Re: FOAF infrastructure changes requested In-Reply-To: <87is9z8vjt.fsf@nwalsh.com> References: <87vfdzkbck.fsf@nwalsh.com> <851c8d3104092708186fb460ea@mail.gmail.com> <87is9z8vjt.fsf@nwalsh.com> Message-ID: <20040927190009.GB15164@homer.w3.org> * Norman Walsh [2004-09-27 13:35-0400] > / Jim Ley was heard to say: > | On Mon, 27 Sep 2004 10:58:19 -0400, Norman Walsh wrote: > |> 1. The FOAF RDF Schema and the FOAF spec are content-negotiated from > |> the namespace URI (http://xmlns.com/foaf/0.1/). This means that a tool > |> like XSLT's document() function cannot get the schema from the > |> namespace URI. > | > | I don't quite see why this follows, surely an XSLT parser must send an > | accept header containing only XML mime-types as acceptable. whilst > | including application/rdf+xml may be too much to expect, ensuring that > | the foaf URI returns XML in response to application/xml aswell should > | be fine. > > Alas, XSLT provides no way to influence the accept headers sent by the > implementation. I imagine that application/xml is sent, but the server > considers the XHTML version a better match. Really? nothing in XSLT 2.0 or EXSLT? [time passes...] [asking Max Froumentin...] Nothing. Bummer. When's XSLT 3.0 planned? > > |> I humbly suggest that the namespace URI should return the schema with > |> a stylesheet PI that does something reasonable and includes a pointer > |> to the human-readable spec. > | > | I really would rather not have XSLT PI's in the foaf spec - to make my > | life easier of course - I also prefer the human readable HTML as the > | default format. Have the best practices people looked at this sort of > | stuff yet? > > Dunno. I'm not sure content negotiating RDF and human text is a good > idea, but I seem to be in the minority. It seemed like a good idea to me too, but then I figured it best to avoid # in namespace URIs for related reasons, and the jury still seems to be out on that one. Dan > _______________________________________________ > rdfweb-dev mailing list > rdfweb-dev@vapours.rdfweb.org > wiki: http://rdfweb.org/topic/FoafProject > http://rdfweb.org/mailman/listinfo/rdfweb-dev From jim.ley at gmail.com Mon Sep 27 19:01:37 2004 From: jim.ley at gmail.com (Jim Ley) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] FOAF infrastructure changes requested In-Reply-To: <41585895.40505@internetalchemy.org> References: <87vfdzkbck.fsf@nwalsh.com> <851c8d3104092708186fb460ea@mail.gmail.com> <41585895.40505@internetalchemy.org> Message-ID: <851c8d310409271201232265b0@mail.gmail.com> On Mon, 27 Sep 2004 19:14:45 +0100, Ian Davis wrote: > This might be problematic since application/xhtml+xml could also be > returned. Both are obviously acceptable to an XSLT parser but they have > differing levels of usefulness :) ah, you also agree application/xhtml+xml is useless too, we just need to make sure danbri doesn't bother returning one of those then... :-) Jim. From iand at internetalchemy.org Mon Sep 27 22:13:49 2004 From: iand at internetalchemy.org (Ian Davis) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] FOAF infrastructure changes requested In-Reply-To: <20040927185653.GA15164@homer.w3.org> References: <87vfdzkbck.fsf@nwalsh.com> <20040927185653.GA15164@homer.w3.org> Message-ID: <4158909D.6070400@internetalchemy.org> On 27/09/2004 19:56, Dan Brickley wrote: > The reason we have an attribute-centric serialization is so the RDF can > be shoved inside XHTML and the content not spill out in downlevel > browsers. For the pure RDF version we could publish an index.rdf which > had an element-based encoding. If someone wants to write the XSLT that > generates one from the other, I could build it into the check scripts. > Ah, I forgot the schema was buried in the xhtml. This should be readable by Norm's XSLT. (Is this the alternative to conneg? just pack it all inside one big XML document :) Ian From ndw at nwalsh.com Tue Sep 28 09:32:10 2004 From: ndw at nwalsh.com (Norman Walsh) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] Re: FOAF infrastructure changes requested In-Reply-To: <20040927190009.GB15164@homer.w3.org> (Dan Brickley's message of "Mon, 27 Sep 2004 15:00:09 -0400") References: <87vfdzkbck.fsf@nwalsh.com> <851c8d3104092708186fb460ea@mail.gmail.com> <87is9z8vjt.fsf@nwalsh.com> <20040927190009.GB15164@homer.w3.org> Message-ID: <87oejq3fj9.fsf@nwalsh.com> / Dan Brickley was heard to say: | * Norman Walsh [2004-09-27 13:35-0400] |> / Jim Ley was heard to say: |> | On Mon, 27 Sep 2004 10:58:19 -0400, Norman Walsh wrote: |> |> 1. The FOAF RDF Schema and the FOAF spec are content-negotiated from |> |> the namespace URI (http://xmlns.com/foaf/0.1/). This means that a tool |> |> like XSLT's document() function cannot get the schema from the |> |> namespace URI. |> | |> | I don't quite see why this follows, surely an XSLT parser must send an |> | accept header containing only XML mime-types as acceptable. whilst |> | including application/rdf+xml may be too much to expect, ensuring that |> | the foaf URI returns XML in response to application/xml aswell should |> | be fine. |> |> Alas, XSLT provides no way to influence the accept headers sent by the |> implementation. I imagine that application/xml is sent, but the server |> considers the XHTML version a better match. | | Really? nothing in XSLT 2.0 or EXSLT? | [time passes...] [asking Max Froumentin...] EXSLT could do it. I suppose it would have made sense to do it for XSLT 2.0, but the document() function is already fairly complex. I think I'll drop it in as a feature request, but at this late stage, I expect to get slapped down pretty firmly. :-/ | It seemed like a good idea to me too, but then I figured it best to | avoid # in namespace URIs for related reasons, and the jury still seems | to be out on that one. Yep. Be seeing you, norm -- Norman Walsh | It is seldom that any liberty is lost http://nwalsh.com/ | all at once.--David Hume -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 188 bytes Desc: not available Url : http://lists.usefulinc.com/pipermail/foaf-dev/attachments/20040928/07a2515d/attachment-0001.pgp From ndw at nwalsh.com Tue Sep 28 09:38:32 2004 From: ndw at nwalsh.com (Norman Walsh) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] Re: FOAF infrastructure changes requested In-Reply-To: <20040927185653.GA15164@homer.w3.org> (Dan Brickley's message of "Mon, 27 Sep 2004 14:56:53 -0400") References: <87vfdzkbck.fsf@nwalsh.com> <20040927185653.GA15164@homer.w3.org> Message-ID: <87k6ue3f8n.fsf@nwalsh.com> / Dan Brickley was heard to say: | * Norman Walsh [2004-09-27 10:58-0400] |> Hello world, |> |> While working to improve the metadata presentation on norman.walsh.name, |> I stumbled over two issues. |> |> 1. The FOAF RDF Schema and the FOAF spec are content-negotiated from |> the namespace URI (http://xmlns.com/foaf/0.1/). This means that a tool |> like XSLT's document() function cannot get the schema from the |> namespace URI. |> |> I humbly suggest that the namespace URI should return the schema with |> a stylesheet PI that does something reasonable and includes a pointer |> to the human-readable spec. | | Do you think XSLT is deployable for such things yet? That's a judgement call, I guess. I publish random bits of XML with stylesheet PIs. Firefox and IE do the right thing. In the long run, I'd like to put a RDDL document at the namespace URI and let resolvers and other lower-level libraries "do the right thing". Of course, that means the APIs I'm using have to allow me to specify the media types I want back. Sigh. | CSS styling can't, | AFAIK, generate hyperlinks, so would be a hypertextual dead-end. Yeah, CSS doesn't work for these applications. | If/when | XSLT is "out there" enough, and browsers default to running untrusted | XSLTs, then it could be the way to go. I'm not sure I would advertise the namespace URI as the URI of the specification so I'd expect fewer people to go off dereferencing the namespace URI in their browsers. |> 2. The spec begins: |> |> dc:title="Friend of a Friend (FOAF) vocabulary" |> dc:description="The Friend of a Friend (FOAF) RDF vocabulary, described using W3C RDF Schema and the Web Ontology Language." |> dc:date="$Date: 2004/09/01 15:37:56 $"> |> ... |> |> I humbly request that it use one-fewer RDF serialization trick :-) and | | The reason we have an attribute-centric serialization is so the RDF can | be shoved inside XHTML and the content not spill out in downlevel | browsers. For the pure RDF version we could publish an index.rdf which | had an element-based encoding. If someone wants to write the XSLT that | generates one from the other, I could build it into the check scripts. Fair enough. A moot point since I can't get the RDF back from the server in my environment. I suppose if I setup a mapping table or something so that I can, I really ought to just load the URIs as RDF graphs anyway and not as XML documents, then it wouldn't matter. Be seeing you, norm -- Norman Walsh | So, are you working on finding that bug http://nwalsh.com/ | now, or are you leaving it until later? | Yes. -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 188 bytes Desc: not available Url : http://lists.usefulinc.com/pipermail/foaf-dev/attachments/20040928/f4b80ac5/attachment-0001.pgp From ndw at nwalsh.com Tue Sep 28 10:03:47 2004 From: ndw at nwalsh.com (Norman Walsh) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] Re: FOAF infrastructure changes requested In-Reply-To: <4158909D.6070400@internetalchemy.org> (Ian Davis's message of "Mon, 27 Sep 2004 23:13:49 +0100") References: <87vfdzkbck.fsf@nwalsh.com> <20040927185653.GA15164@homer.w3.org> <4158909D.6070400@internetalchemy.org> Message-ID: <87fz523e2k.fsf@nwalsh.com> / Ian Davis was heard to say: | On 27/09/2004 19:56, Dan Brickley wrote: | |> The reason we have an attribute-centric serialization is so the RDF can |> be shoved inside XHTML and the content not spill out in downlevel |> browsers. For the pure RDF version we could publish an index.rdf which |> had an element-based encoding. If someone wants to write the XSLT that |> generates one from the other, I could build it into the check scripts. |> | Ah, I forgot the schema was buried in the xhtml. This should be | readable by Norm's XSLT. Yeah, I suppose. I hadn't noticed it was stuck on the end. Be seeing you, norm -- Norman Walsh | If you are losing your leisure, look http://nwalsh.com/ | out! You may be losing your | soul.--Logan Pearsall Smith -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 188 bytes Desc: not available Url : http://lists.usefulinc.com/pipermail/foaf-dev/attachments/20040928/0647310d/attachment-0001.pgp From Libby.Miller at bristol.ac.uk Tue Sep 28 19:10:33 2004 From: Libby.Miller at bristol.ac.uk (Libby Miller) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] action: date of birth proposals Message-ID: ACTION libby: send mail to rdfweb-dev outlining 3 proposals for date or birth / birthday http://rdfig.xmlhack.com/2004/09/22/2004-09-22.html#1095870789.971234 Log: http://www.ilrt.bris.ac.uk/discovery/chatlogs/rdfig/2004-09-22.html#T17-32-02 Summary: (with snips) [[ 17:32:06 As I recall it...: 17:32:22 - a few old foaf files, like mine, have a plain literal field called dateOfBirth 17:32:28 that takes things like 1972-01-09 17:32:54 ...but it never got in spec, cos we also wanted to be able to say 'my birthday is jan 9th', without giving away age 17:33:11 useful for orkut-style birthday reminders, and a bit less revealing (some security issues re identity theft potential) ]] Three proposals: 1. Two new properties: foaf:dateOfBirth and foaf:birthday * dateOfBirth could be a string, datatype or object (non Gregorian calendars?), perhaps of the form 2004-09-22 (using http://www.w3.org/TR/xmlschema-2/#date) * similar considerations apply to birthday, except it would be of teh form 09-22 (MM-DD) (using http://www.w3.org/TR/xmlschema-2/#gMonthDay) 2. Three separate properties: dayOfBirth, monthOfBirth, yearOfBirth * similar considerations apply as above: http://www.w3.org/TR/xmlschema-2/#gYear http://www.w3.org/TR/xmlschema-2/#gMonth http://www.w3.org/TR/xmlschema-2/#gDay I'm not totally convinced the xml schema datatypes work quite correctly there, e.g. [[ gDay is a gregorian day that recurs, specifically a day of the month such as the 5th of the month ]] 3. Use existing bio:event structure bio vocab: http://vocab.org/bio/0.1/ - example on the front page: [[ 1970-06-15 Brentwood, Essex, United Kingdom ]] I'm not sure quite how birthday would work like this. Essentially it's a recurring event (which I tried to model in ical a while back: http://rdfweb.org/pipermail/rdfweb-dev/2003-November/012191.html) An additional idea might just be to use rdf ical: 4. Use RDF ical --the advantage here might be that you could download birthdays to PIMS. any thoughts/preferences? Libby From bkdelong at pobox.com Tue Sep 28 19:18:07 2004 From: bkdelong at pobox.com (B.K. DeLong) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] action: date of birth proposals In-Reply-To: Message-ID: <5.2.1.1.2.20040928151429.02c8f080@mail.brain-stream.com> At 08:10 PM 9/28/2004 +0100, Libby Miller wrote: >2. Three separate properties: dayOfBirth, monthOfBirth, yearOfBirth > * similar considerations apply as above: >http://www.w3.org/TR/xmlschema-2/#gYear >http://www.w3.org/TR/xmlschema-2/#gMonth >http://www.w3.org/TR/xmlschema-2/#gDay > >I'm not totally convinced the xml schema datatypes work quite correctly >there, e.g. > >[[ >gDay is a gregorian day that recurs, specifically a day of the month such >as the 5th of the month >]] I think 3 is the way to go. As encryption and control of data becomes increasingly important in the FOAF community, people need the ability to be able to specify just how much of their "birthday equation" someone can have via trust-relationships. By separating it out into 3 different properties, an application could spit out a single property, a pair, or all three. >3. Use existing bio:event structure > >bio vocab: http://vocab.org/bio/0.1/ >- example on the front page: >[[ > > > > 1970-06-15 > Brentwood, Essex, United Kingdom > > > >]] > >I'm not sure quite how birthday would work like this. Essentially it's a >recurring event (which I tried to model in ical a while back: >http://rdfweb.org/pipermail/rdfweb-dev/2003-November/012191.html) If we use Bio, I'd like to see the date broken out into 3 parts. BTW, one thing we are missing - timeofBirth ;) It only really makes a real difference to twins and astrologers but it is a dataset we have nowadays. -- B.K. DeLong bkdelong@pobox.com +1.617.797.2472 http://bkdelong.mit.edu Work. http://www.brain-stream.com Play. http://www.the-leaky-cauldron.org Potter. http://www.city-of-doors.com Sigil. http://www.hackerfoundation.org Future. http://www.osvdb.org/ Security. http://www.wkdelong.org Son. PGP Fingerprint: 38D4 D4D4 5819 8667 DFD5 A62D AF61 15FF 297D 67FE FOAF: http://foaf.brain-stream.org From bkdelong at pobox.com Tue Sep 28 19:20:52 2004 From: bkdelong at pobox.com (B.K. DeLong) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] action: date of birth proposals In-Reply-To: <5.2.1.1.2.20040928151429.02c8f080@mail.brain-stream.com> References: Message-ID: <5.2.1.1.2.20040928152028.02c825e8@mail.brain-stream.com> At 03:18 PM 9/28/2004 -0400, B.K. DeLong wrote: >I think 3 is the way to go. As encryption and control of data becomes >increasingly important in the FOAF community, people need the ability to >be able to specify just how much of their "birthday equation" someone can >have via trust-relationships. By separating it out into 3 different >properties, an application could spit out a single property, a pair, or >all three. Blast. I meant #2. -- B.K. DeLong bkdelong@pobox.com +1.617.797.2472 http://bkdelong.mit.edu Work. http://www.brain-stream.com Play. http://www.the-leaky-cauldron.org Potter. http://www.city-of-doors.com Sigil. http://www.hackerfoundation.org Future. http://www.osvdb.org/ Security. http://www.wkdelong.org Son. PGP Fingerprint: 38D4 D4D4 5819 8667 DFD5 A62D AF61 15FF 297D 67FE FOAF: http://foaf.brain-stream.org From jim at jibbering.com Tue Sep 28 19:30:23 2004 From: jim at jibbering.com (Jim Ley) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] Re: action: date of birth proposals References: Message-ID: "Libby Miller" wrote in message news:Pine.GSO.4.61.0409281947450.6099@mail.ilrt.bris.ac.uk... > 2. Three separate properties: dayOfBirth, monthOfBirth, yearOfBirth > * similar considerations apply as above: > http://www.w3.org/TR/xmlschema-2/#gYear > http://www.w3.org/TR/xmlschema-2/#gMonth > http://www.w3.org/TR/xmlschema-2/#gDay I like this, it allows you to partial information, rather than being forced to give out everything. Jim. From Libby.Miller at bristol.ac.uk Tue Sep 28 19:38:15 2004 From: Libby.Miller at bristol.ac.uk (Libby Miller) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] proposed agenda foaf meeting 2004-09-29, 1630 UTC Message-ID: hi all, The next Foaf project meeting is at 1630 UTC 2004-09-29 for 60-90 minutes. In your timezone: http://www.timeanddate.com/worldclock/fixedtime.html?day=29&month=9&year=2004&hour=16&min=30&sec=0&p1=0 I propose we use irc.freenode.net #rdfig again rather than #foaf (for the tools) Last meeting: Agenda, actions: http://rdfig.xmlhack.com/2004/09/22/2004-09-22.html logs: http://www.ilrt.bris.ac.uk/discovery/chatlogs/rdfig/2004-09-22.html#T16-28-52 Actions from last meeting: * danbri update FOAF spec to add a main section on naming, to provide a unified treatment of first/last/given/family addressing Misha's concern re first/last - continued * action: crschmidt to circulate crawler stats on foaf:nick foaf:name foaf:firstName foaf:givenname foaf:surname foaf:family_name - continued * ACTION libby: send mail to rdfweb-dev outlining 3 proposals for date or birth / birthday - see http://rdfweb.org/pipermail/rdfweb-dev/2004-September/013733.html * ACTION danbri: invite Peter St Andre to a future IRC FOAF meeting * ACTION danbri propose to the list "that it should be a goal to ultimately get the FOAF spec in a shape (whatever shape that is) such that IETF, Dublin Core and W3C Working Groups are comfortable citing it when they need to draw upon core concepts like Person, homepage, mbox, mbox_sha1sum etc." * action danbri: get his FOAF talks from FOAFCamp and Galway online (and do the workshop report!) * action: danbri let nicole know that Jim had problems with fontsize on new sit * ACTION bengee propose some clarifications re lifecycle/stability vocab to list, goal of having better machine-readable status for FOAF namespace - see http://rdfweb.org/pipermail/rdfweb-dev/2004-September/013705.html Agenda proposals for 2004-09-29 meeting * Domain of foaf:interest - Morten http://rdfweb.org/pipermail/rdfweb-dev/2004-September/013708.html + thread * Foaf views and profiles 17:32:27 pjz: views or profiles could probably be made another action/topic for the regular irc meetings? 17:34:10 * teefal seconds profiles/views as future topic http://www.ilrt.bris.ac.uk/discovery/chatlogs/rdfig/2004-09-22.html#T17-32-27 and possibly... * Norm's mail http://rdfweb.org/pipermail/rdfweb-dev/2004-September/013718.html More welcome! Libby From stpeter at jabber.org Tue Sep 28 20:29:03 2004 From: stpeter at jabber.org (Peter Saint-Andre) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] Re: proposed agenda foaf meeting 2004-09-29, 1630 UTC References: Message-ID: In article , Libby Miller wrote: > * ACTION danbri: invite Peter St Andre to a future IRC FOAF meeting Wow, I have my very own action item! ;-) I'll make every attempt to be there tomorrow. /psa From stpeter at jabber.org Tue Sep 28 20:54:33 2004 From: stpeter at jabber.org (Peter Saint-Andre) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] Re: action: date of birth proposals References: Message-ID: In article , Libby Miller wrote: > 2. Three separate properties: dayOfBirth, monthOfBirth, yearOfBirth > * similar considerations apply as above: > http://www.w3.org/TR/xmlschema-2/#gYear > http://www.w3.org/TR/xmlschema-2/#gMonth > http://www.w3.org/TR/xmlschema-2/#gDay This seems best (clearest and most flexible) to me as well. /psa From richard at cyganiak.de Tue Sep 28 22:10:00 2004 From: richard at cyganiak.de (Richard Cyganiak) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] action: date of birth proposals In-Reply-To: References: Message-ID: <23DBFCAA-119B-11D9-99BF-000A95DABFDC@cyganiak.de> > ACTION libby: send mail to rdfweb-dev outlining 3 proposals for date > or birth / birthday ... > 1. Two new properties: foaf:dateOfBirth and foaf:birthday I like foaf:birthday because it lets you trivially find people with the same birthday (as in same month and day). But why not avoid the overlap? foaf:birthday, e.g. "07-15"^^xsd:gDayMonth foaf:yearOfBirth, e.g. "1979"^^xsd:gYear The year could still be omitted, and the reason for there being two properties would be clearer. Also smaller risk of accidently confusing the two props. I also like that both properties are useful in isolation. Birthday tells you when to congratulate. YearOfBirth tells you the age. Very natural. Splitting birthday into dayOfBirth and monthOfBirth, like in proposal 2, gives you two quite awkward properties that are only useful if both are known. Richard From danbri at w3.org Tue Sep 28 22:49:16 2004 From: danbri at w3.org (Dan Brickley) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] Re: proposed agenda foaf meeting 2004-09-29, 1630 UTC In-Reply-To: References: Message-ID: <20040928224916.GF9755@homer.w3.org> * Peter Saint-Andre [2004-09-28 14:29-0600] > In article , > Libby Miller > wrote: > > > * ACTION danbri: invite Peter St Andre to a future IRC FOAF meeting > > Wow, I have my very own action item! ;-) > > I'll make every attempt to be there tomorrow. That'd be great. I declare my ACTION completed ;) Would you be able to lead 5-10 mins discussion on FOAF-related requirements/issues etc in a Jabber/XMPP context? You've previously mentioned concerns about imposing a requirement for an RDF parser on lightweight Jabber clients. I hope we can find a way to distinguish (i) having an RDF representation of Jabber profile data from (ii) having that encoded directly in RDF/XML, eg. http://www.w3.org/TR/2004/NOTE-grddl-20040413/ suggests one (XSLT-based) technique for mapping non-RDF/XML markup idioms into the RDF data model. If you could give us a sense of where things are up to in the Jabber world, and the kinds of profile information you'd be interested to carry, I think that'd be really helpful. Anything that you could send as email'd background reading would of course be much appreciated. cheers, Dan From stpeter at jabber.org Tue Sep 28 23:05:22 2004 From: stpeter at jabber.org (Peter Saint-Andre) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] Re: proposed agenda foaf meeting 2004-09-29, 1630 UTC In-Reply-To: <20040928224916.GF9755@homer.w3.org> References: <20040928224916.GF9755@homer.w3.org> Message-ID: On Sep 28, 2004, at 4:49 PM, Dan Brickley wrote: > Would you be able to lead 5-10 mins discussion on FOAF-related > requirements/issues etc in a Jabber/XMPP context? I'd be happy to. > You've previously mentioned concerns about imposing a requirement for > an > RDF parser on lightweight Jabber clients. I hope we can find a way to > distinguish (i) having an RDF representation of Jabber profile data > from > (ii) having that encoded directly in RDF/XML, eg. > http://www.w3.org/TR/2004/NOTE-grddl-20040413/ suggests one > (XSLT-based) > technique for mapping non-RDF/XML markup idioms into the RDF data > model. There's been some discussion about this on the Standards-JIG list, which is where Jabber protocol talk happens (see this post and follow-ups if you're interested): http://mail.jabber.org/pipermail/standards-jig/2004-September/ 006254.html > If you could give us a sense of where things are up to in the Jabber > world, and the kinds of profile information you'd be interested to > carry, I think that'd be really helpful. Anything that you could send > as > email'd background reading would of course be much appreciated. Probably the best background info right now is in my blog: http://www.saint-andre.com/blog/2004-09.html#2004-09-17T09:25 I hope to write up something more formal (i.e., a JEP) in a week or two. Peter From am2stewa at uwaterloo.ca Wed Sep 29 02:32:30 2004 From: am2stewa at uwaterloo.ca (Alex Stewart) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] Does rdfs:seeAlso imply anything? Message-ID: <20040929023230.GC334@scg.math> Hi, If we have blank1 rdfs:seeAlso uri1 and uri1 foaf:primaryTopic blank2 then can blank1 and blank2 be smushed into one node? I've searched the archive for "seeAlso topic" and I've found a message about PPDs from Morten[1] that suggests no, but then I've also seen the second mode of thought in PointingFromPersonToGroup in the wiki[2] that hints yes (since the foaf:Group doesn't have any inverse functional properties). Another yes vote comes from Leigh's Introduction to FOAF[3] (search for "Harry"; it's right after the first occurance). Mind you, people aren't going out and saying "yes" to this question, but rather it's just implied in the RDF statements that they write. Alex 1. http://rdfweb.org/pipermail/rdfweb-dev/2004-February/012663.html 2. http://rdfweb.org/topic/PointingFromPersonToGroup 3. http://www.xml.com/pub/a/2004/02/04/foaf.html From mof-rdf at mfd-consult.dk Wed Sep 29 06:06:47 2004 From: mof-rdf at mfd-consult.dk (Morten Frederiksen) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] Does rdfs:seeAlso imply anything? In-Reply-To: <20040929023230.GC334@scg.math> References: <20040929023230.GC334@scg.math> Message-ID: <200409290806.47988.mof-rdf@mfd-consult.dk> Hi Alex, On Wednesday 29 September 2004 04:32, Alex Stewart wrote: > If we have > blank1 rdfs:seeAlso uri1 > and > uri1 foaf:primaryTopic blank2 > then can blank1 and blank2 be smushed into one node? I stand firmly by my view as you've pointed at, the answer is no. > I've searched the archive for "seeAlso topic" and I've found a message > about PPDs from Morten[1] that suggests no, but then I've also seen the > second mode of thought in PointingFromPersonToGroup in the wiki[2] that > hints yes (since the foaf:Group doesn't have any inverse functional > properties). I think you're reading too much into that. The connection between the group and the person in that example is formed by the use of the foaf:member property, not the rdfs:seeAlso, note the comment, "ME, IDENTIFIED IN THE NODEID ABOVE", which is hinting at something like: ... ... > Another yes vote comes from Leigh's Introduction to > FOAF[3] (search for "Harry"; it's right after the first occurance). Hmm, while I do see why it can be seen as a "yes vote", I think it's simply a case of the example missing identifying properties for all the person's mentioned. > Mind you, people aren't going out and saying "yes" to this question, but > rather it's just implied in the RDF statements that they write. Agreed, in many cases it is indeed true, but changing the semantics of rdfs:seeAlso (or overloading it) isn't needed, as in those cases some (other) identifying properties are usually used as well. Regards, Morten From mof-rdf at mfd-consult.dk Wed Sep 29 06:19:38 2004 From: mof-rdf at mfd-consult.dk (Morten Frederiksen) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] foaf:interest et al. In-Reply-To: <41581515.4060601@internetalchemy.org> References: <41581515.4060601@internetalchemy.org> Message-ID: <200409290819.38559.mof-rdf@mfd-consult.dk> Hi, On Monday 27 September 2004 15:26, Ian Davis wrote: > I've been having a think about foaf:interest and its close buddies > foaf:topic and foaf:interest_topic which I've written about here[1] Very nice writeup. > I think the foaf:interest property is misnamed. I really don't disagree on that one, but the definition is solid and it's being used, which I think are two more important points. > My suggestion is to rename the foaf:topic_interest property to > foaf:topicOfInterest or foaf:interestInTopic with the definition: A > topic that this person expresses an interest in. Sounds quite reasonable, even if the topic is hard to define, see below. > Then, for completeness > I'd also suggest the introduction of a foaf:Topic class. The domain of > foaf:topicOfInterest and foaf:topic would be foaf:Topic. The range of > foaf:topicOfInterest would be foaf:Agent. Hmm, I think we should think (at least) twice about this, since you'll soon realise that practially anything can be a topic: _:me foaf:interest . If we agree that the above is a reasonable statement, then that'd make His Bobness a foaf:Topic, which might not be wrong, but seems somewhat strange. Exactly what is a "topic"? Perhaps some SKOS [1] stuff could be pulled in here? [1] http://www.w3.org/2001/sw/Europe/reports/thes/1.0/guide/ Regards, Morten From tony-rdf at kasei.com Wed Sep 29 08:26:54 2004 From: tony-rdf at kasei.com (Tony Bowden) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] action: date of birth proposals In-Reply-To: <23DBFCAA-119B-11D9-99BF-000A95DABFDC@cyganiak.de> References: <23DBFCAA-119B-11D9-99BF-000A95DABFDC@cyganiak.de> Message-ID: <20040929082654.GA26944@soto.kasei.com> On Wed, Sep 29, 2004 at 12:10:00AM +0200, Richard Cyganiak wrote: > I also like that both properties are useful in isolation. Birthday > tells you when to congratulate. YearOfBirth tells you the age. Very > natural. Erm? You can't tell age just from year of birth ... Tony From danbri at w3.org Wed Sep 29 09:05:00 2004 From: danbri at w3.org (Dan Brickley) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] Does rdfs:seeAlso imply anything? In-Reply-To: <20040929023230.GC334@scg.math> References: <20040929023230.GC334@scg.math> Message-ID: <20040929090500.GA6822@homer.w3.org> Hi Alex, * Alex Stewart [2004-09-28 22:32-0400] > Hi, > > If we have > > blank1 rdfs:seeAlso uri1 > and > uri1 foaf:primaryTopic blank2 > > then can blank1 and blank2 be smushed into one node? > > I've searched the archive for "seeAlso topic" and I've found a message > about PPDs from Morten[1] that suggests no, but then I've also seen the > second mode of thought in PointingFromPersonToGroup in the wiki[2] that > hints yes (since the foaf:Group doesn't have any inverse functional > properties). Another yes vote comes from Leigh's Introduction to > FOAF[3] (search for "Harry"; it's right after the first occurance). > > Mind you, people aren't going out and saying "yes" to this question, but > rather it's just implied in the RDF statements that they write. I think the answer here is 'no'; a resource can be the rdfs:seeAlso of something without necessarily being its foaf:primaryTopic. An example might help, danbri libby A document about libby that mentions danbri in passing blank1 is a node that stands for me. We are truly told that, w.r.t. me, we can seealso a.rdf for maybe some more information. This is pretty weak, weaker than either foaf:primaryTopic or foaf:topic. But it's true and useful. The example has a.rdf being mostly about someone else, with a passing mention of me (eg. a co-depiction, or co-authorship). In such cases it's often a judgement call whether to use foaf:topic vs foaf:primaryTopic. I've mentioned before here that deferring to the document's own claims about it's (primary)topic is probably a good strategy. http://esw.w3.org/topic/UsingSeeAlso has some more notes on rdfs:seeAlso. Hope this helps, Dan > Alex > > 1. http://rdfweb.org/pipermail/rdfweb-dev/2004-February/012663.html > 2. http://rdfweb.org/topic/PointingFromPersonToGroup > 3. http://www.xml.com/pub/a/2004/02/04/foaf.html > > _______________________________________________ > rdfweb-dev mailing list > rdfweb-dev@vapours.rdfweb.org > wiki: http://rdfweb.org/topic/FoafProject > http://rdfweb.org/mailman/listinfo/rdfweb-dev From GK at ninebynine.org Wed Sep 29 07:31:27 2004 From: GK at ninebynine.org (Graham Klyne) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] action: date of birth proposals In-Reply-To: Message-ID: <5.1.0.14.2.20040929082357.00bbf4e0@127.0.0.1> Considering in the abstract (i.e. not considering practical details) I'd prefer something based on RDF iCal, on the basis that will (hopefully) help to encourage uniform representation (using RDF iCal) of dates across different RDF applications. The practical concern is that setting up an Ical recurring event structure for such a simple piece of information may be overly complicated for such a simple task, which suggests some kind of simple date-as-literal may be desired. I note that ISO 8601 (upon which I understand the XML schema date type, and RFC 3339, are based) allows a string of the form: --MMDD or --MM-DD for month+day only. (See ISO 8601:1988, section 5.2.1.3) #g -- At 20:10 28/09/04 +0100, Libby Miller wrote: >ACTION libby: send mail to rdfweb-dev outlining 3 proposals for date or >birth / birthday >http://rdfig.xmlhack.com/2004/09/22/2004-09-22.html#1095870789.971234 > >Log: >http://www.ilrt.bris.ac.uk/discovery/chatlogs/rdfig/2004-09-22.html#T17-32-02 > >Summary: (with snips) >[[ >17:32:06 As I recall it...: >17:32:22 - a few old foaf files, like mine, have a plain literal >field called dateOfBirth >17:32:28 that takes things like 1972-01-09 >17:32:54 ...but it never got in spec, cos we also wanted to be >able to say 'my birthday is jan 9th', without giving away age >17:33:11 useful for orkut-style birthday reminders, and a bit >less revealing (some security issues re identity theft potential) >]] > >Three proposals: > >1. Two new properties: foaf:dateOfBirth and foaf:birthday > * dateOfBirth could be a string, datatype or object (non Gregorian > calendars?), perhaps of the form 2004-09-22 (using > http://www.w3.org/TR/xmlschema-2/#date) > * similar considerations apply to birthday, except it would be of teh > form 09-22 (MM-DD) (using http://www.w3.org/TR/xmlschema-2/#gMonthDay) > >2. Three separate properties: dayOfBirth, monthOfBirth, yearOfBirth > * similar considerations apply as above: >http://www.w3.org/TR/xmlschema-2/#gYear >http://www.w3.org/TR/xmlschema-2/#gMonth >http://www.w3.org/TR/xmlschema-2/#gDay > >I'm not totally convinced the xml schema datatypes work quite correctly >there, e.g. > >[[ >gDay is a gregorian day that recurs, specifically a day of the month such >as the 5th of the month >]] > >3. Use existing bio:event structure > >bio vocab: http://vocab.org/bio/0.1/ >- example on the front page: >[[ > > > > 1970-06-15 > Brentwood, Essex, United Kingdom > > > >]] > >I'm not sure quite how birthday would work like this. Essentially it's a >recurring event (which I tried to model in ical a while back: >http://rdfweb.org/pipermail/rdfweb-dev/2003-November/012191.html) > >An additional idea might just be to use rdf ical: > >4. Use RDF ical > >--the advantage here might be that you could download birthdays to PIMS. > > >any thoughts/preferences? > >Libby > >_______________________________________________ >rdfweb-dev mailing list >rdfweb-dev@vapours.rdfweb.org >wiki: http://rdfweb.org/topic/FoafProject >http://rdfweb.org/mailman/listinfo/rdfweb-dev ------------ Graham Klyne For email: http://www.ninebynine.org/#Contact From bnowack at appmosphere.com Wed Sep 29 09:35:24 2004 From: bnowack at appmosphere.com (Benjamin Nowack) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] action: date of birth proposals In-Reply-To: <20040929082654.GA26944@soto.kasei.com> References: <23DBFCAA-119B-11D9-99BF-000A95DABFDC@cyganiak.de> <20040929082654.GA26944@soto.kasei.com> Message-ID: On 29.09.2004 09:26:54, Tony Bowden wrote: >On Wed, Sep 29, 2004 at 12:10:00AM +0200, Richard Cyganiak wrote: >> I also like that both properties are useful in isolation. Birthday >> tells you when to congratulate. YearOfBirth tells you the age. Very >> natural. > >Erm? > >You can't tell age just from year of birth ... But I'd say you can tell it in a "sufficiently precise" way for the majority of related use cases. If I see someone with a yearOfBirth=1973, wouldn't that be sufficiently precise for saying that we have the same age, no matter if (s)he is still 30 or already 31? benjamin -- Benjamin Nowack Kruppstr. 100 45145 Essen, Germany http://www.appmosphere.com/ > >Tony > >_______________________________________________ >rdfweb-dev mailing list >rdfweb-dev@vapours.rdfweb.org >wiki: http://rdfweb.org/topic/FoafProject >http://rdfweb.org/mailman/listinfo/rdfweb-dev > > From bnowack at appmosphere.com Wed Sep 29 09:59:35 2004 From: bnowack at appmosphere.com (Benjamin Nowack) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] action: date of birth proposals In-Reply-To: <5.1.0.14.2.20040929082357.00bbf4e0@127.0.0.1> References: <5.1.0.14.2.20040929082357.00bbf4e0@127.0.0.1> Message-ID: On 29.09.2004 08:31:27, Graham Klyne wrote: >Considering in the abstract (i.e. not considering practical details) I'd >prefer something based on RDF iCal, on the basis that will (hopefully) help >to encourage uniform representation (using RDF iCal) of dates across >different RDF applications. > >The practical concern is that setting up an Ical recurring event structure >for such a simple piece of information may be overly complicated for such a >simple task, which suggests some kind of simple date-as-literal may be >desired. I note that ISO 8601 (upon which I understand the XML schema date >type, and RFC 3339, are based) allows a string of the form: > --MMDD >or > --MM-DD >for month+day only. (See ISO 8601:1988, section 5.2.1.3) there were two ideas behind proposing three separate props. one is the flexibility of what people might want to say about themselves, which could be solved by a single (e.g. --mm-dd / yyyy-mm-dd) or two properties (birthday, yearOfBirth) as people are probably not going to tell their dayOfBirth without the monthOfBirth. but the other objective was the ease of querying simple triple stores. (or even standard ones as long as we can't be sure that regexps/date handling will be part of the DAWG recommendation). having separate props should make it possible to query even the simplest triple table for e.g.: - people who have their birthday this/next month (via the month prop) - people with a birthday for a given month-day-combination (via day/month) - people of a certain age (via year) but I admit, this is concerning only a limited practical view. RDF iCal compatibility could be another (maybe more) important one. benjamin -- Benjamin Nowack Kruppstr. 100 45145 Essen, Germany http://www.appmosphere.com/ > >#g >-- > >------------ >Graham Klyne >For email: >http://www.ninebynine.org/#Contact From ldodds at ingenta.com Wed Sep 29 12:41:00 2004 From: ldodds at ingenta.com (Leigh Dodds) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] Does rdfs:seeAlso imply anything? In-Reply-To: <200409290806.47988.mof-rdf@mfd-consult.dk> References: <20040929023230.GC334@scg.math> <200409290806.47988.mof-rdf@mfd-consult.dk> Message-ID: <415AAD5C.1020902@ingenta.com> Morten Frederiksen wrote: >>Another yes vote comes from Leigh's Introduction to >>FOAF[3] (search for "Harry"; it's right after the first occurance). > > Hmm, while I do see why it can be seen as a "yes vote", I think it's simply a > case of the example missing identifying properties for all the person's > mentioned. Yes, it shouldn't be taken as a "yes vote", it was just meant to be a concise example showing use of foaf:knows and rdfs:seeAlso linking, you shouldn't read anything else into it. L. From iand at internetalchemy.org Wed Sep 29 12:43:20 2004 From: iand at internetalchemy.org (Ian Davis) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] foaf:interest et al. In-Reply-To: <200409290819.38559.mof-rdf@mfd-consult.dk> References: <41581515.4060601@internetalchemy.org> <200409290819.38559.mof-rdf@mfd-consult.dk> Message-ID: <415AADE8.4090809@internetalchemy.org> On 29/09/2004 07:19, Morten Frederiksen wrote: >>I've been having a think about foaf:interest and its close buddies >>foaf:topic and foaf:interest_topic which I've written about here[1] > > Very nice writeup. Thank you >>I think the foaf:interest property is misnamed. > > I really don't disagree on that one, but the definition is solid and it's > being used, which I think are two more important points. I accept both those points but I think there is a fundamental issue at stake here. Why do we use human readable terms in data designed foer machine reading? The answer, of course, is that being human readable has many advantages in terms of authoring and validation. The names we give to vocabulary terms are important. They serve as mnemonics so we don't have to refer to the specification whenever we write some RDF. They also draw on shared context which makes them easier to use and easier to understand. The problem I see with foaf:interest is that the name given to it has a very general pre-existing meaning, whereas the FOAF definition is very specific. This is in contrast to foaf:knows which is a general term with a general definition, or foaf:jabberID with is very specific. >>Then, for completeness >>I'd also suggest the introduction of a foaf:Topic class. The domain of >>foaf:topicOfInterest and foaf:topic would be foaf:Topic. The range of >>foaf:topicOfInterest would be foaf:Agent. I completely messed this up. I meant: The range of foaf:topicOfInterest and foaf:topic would be foaf:Topic The domain of foaf:topicOfInterest would be foaf:Agent In that: [a foaf:Agent] foaf:topicOfInterest [a foaf:Topic] . > > Hmm, I think we should think (at least) twice about this, since you'll soon > realise that practially anything can be a topic: > > _:me foaf:interest . > > If we agree that the above is a reasonable statement, then that'd make His > Bobness a foaf:Topic, which might not be wrong, but seems somewhat strange. I think it's quite appropriate to consider Bob Dylan as a person a topic of interest. His music, poetry, lifestyle, philosophy, smoking habits and fan demographics are also valid topics. > Exactly what is a "topic"? I can be flippant and say it's the range of foaf:topic. More seriously, WordNet sense 2 is closest "some situation or event that is thought about" http://www.cogsci.princeton.edu/cgi-bin/webwn2.0?stage=1&word=topic > Perhaps some SKOS [1] stuff could be pulled in here? Probably. I need to read it instead of skim it. Ian From iand at internetalchemy.org Wed Sep 29 12:48:37 2004 From: iand at internetalchemy.org (Ian Davis) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] proposed agenda foaf meeting 2004-09-29, 1630 UTC In-Reply-To: References: Message-ID: <415AAF25.8010405@internetalchemy.org> I fully intended to make this meeting but can't because of a family situation. I'm feeling bad because it was me who specifically requested a late afternoon meeting at Galway and I haven't been able to attend one yet! Please accept my apologies and hopefully I'll make the next one. Ian From danbri at w3.org Wed Sep 29 14:12:00 2004 From: danbri at w3.org (Dan Brickley) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] a FOAF goal: spec fit for citation from standards docs Message-ID: <20040929141158.GI2012@homer.w3.org> In last IRC meeting, http://rdfig.xmlhack.com/2004/09/22/2004-09-22.html I said I'd drop the FOAF list a note proposing... "that it should be a goal to ultimately get the FOAF spec in a shape (whatever shape that is) such that IETF, Dublin Core and W3C Working Groups are comfortable citing it when they need to draw upon core concepts like Person, homepage, mbox, mbox_sha1sum etc." This is pretty broad-brush, necessarily, since these various organizations have different --- and evolving --- processes and conventions for citing externally managed specs. Without getting into distinction between normative and informational citations, I'd like to get a sense of whether folk here think this is generally a healthy goal. I do. It seems a shame to do all this work, only to have standards-track efforts feel that they'd need to duplicate it in a more formal setting. At least if bits of it do get duplicated/shadowed/etc., we should, I think, get FOAF into a form such that the editors of these specs feel comfortable making prose and RDF/XML references to the terms defined in the FOAF spec. Guess that means I need to take out the nudie pics from the spec... Dan From danbri at w3.org Wed Sep 29 14:13:08 2004 From: danbri at w3.org (Dan Brickley) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] proposed agenda foaf meeting 2004-09-29, 1630 UTC In-Reply-To: References: Message-ID: <20040929141308.GJ2012@homer.w3.org> * Libby Miller [2004-09-28 20:38+0100] > > hi all, > > * ACTION danbri: invite Peter St Andre to a future IRC FOAF meeting Done; welcome, peter. > * ACTION danbri propose to the list "that it should be a goal to > ultimately get the FOAF spec in a shape (whatever shape that is) such > that IETF, Dublin Core and W3C Working Groups are comfortable citing it > when they need to draw upon core concepts like Person, homepage, mbox, > mbox_sha1sum etc." Done > * action danbri: get his FOAF talks from FOAFCamp and Galway online (and > do the workshop report!) In progress (aka not done); continued. > * action: danbri let nicole know that Jim had problems with fontsize on > new sit Done; From iand at internetalchemy.org Wed Sep 29 16:07:53 2004 From: iand at internetalchemy.org (Ian Davis) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] a FOAF goal: spec fit for citation from standards docs In-Reply-To: <20040929141158.GI2012@homer.w3.org> References: <20040929141158.GI2012@homer.w3.org> Message-ID: <415ADDD9.4030704@internetalchemy.org> On 29/09/2004 15:12, Dan Brickley wrote: > In last IRC meeting, http://rdfig.xmlhack.com/2004/09/22/2004-09-22.html > I said I'd drop the FOAF list a note proposing... > > "that it should be a goal to ultimately get the FOAF spec in a shape > (whatever shape that is) such that IETF, Dublin Core and W3C Working Groups > are comfortable citing it when they need to draw upon core concepts > like Person, homepage, mbox, mbox_sha1sum etc." > > ... I'd like to > get a sense of whether folk here think this is generally a healthy goal. I also think this is a goal worth striving for. It doesn't have to mean a full-on standards process, but it does mean having a well-defined change process (which I think we are close to having). It also means we need to make certain guarantees about stability (again, this is being worked on). Being slightly tangential for a moment, I just want to re-float my suggestion that new terms be introduced in a separate namespace until they are mature enough to move into the core. I suggested this at Galway although I'm not sure I articulated it very clearly. My proposal is to create a new namespace http://xmlns.com/foaf/experimental/ into which new experimental terms are put. Implementors are made aware that terms in this namespace may change meaning, subdivide or disappear completely with little or no notice. Once a term has been around for long enough for us to be confident we understand how it fits into FOAF and its usefulness then it can migrate into the core namespace. The key advantage of this approach is that the creativity of the FOAF community isn't stifled by the requirements to be stable and serious about the spec. There are a number of disadvantages as you pointed out in Galway, such applications relying on particular namespace strings. Ian From Libby.Miller at bristol.ac.uk Thu Sep 30 13:49:12 2004 From: Libby.Miller at bristol.ac.uk (Libby Miller) Date: Mon Dec 18 17:23:57 2006 Subject: [rdfweb-dev] next foaf meeting? Message-ID: Dan's not around next wednesday, but I'm happy to hold a foaf meeting if there's demand. It would be 1630 UTC, 6th October, on #rdfig. Ping me offlist if you think you can make it and I'll send something in the next couple of days yes or no. Agenda items welcome too, preferably to the list. In your timezone: http://www.timeanddate.com/worldclock/fixedtime.html?day=6&month=10&year=2004&hour=16&min=30&sec=0&p1=0 cheers, Libby