[foaf-dev] Discovery, ID and FOAF

[Peter Williams]

In 6 months of starting from scratch with semweb ideas ... in order to expe=
riment with what folks other than ourselves are practically doing with open=
id, the nearest 2 things Ive seen to Henry's original ideas are:

1. at http://myopenid.com/, not only can the service send out an openid ass=
ertion to those who would rely on your id, your contacts can even just acce=
ss "your" page at myopenid. The page has semantic markup in the class tags =
of the XHTML, using vcard tags. The tagged attributes about you (first name=
, last name, etc) are the same values as the openid sever is willing to ass=
ert to consuming sites. The whole point of openid is that the process of de=
liverring one or more of those attributes to any particular using website i=
s controlled in realtime by you, providing a kind of "personal access contr=
ol" model. It easier if one drops the notion of thinking in terms of acess =
control, and one shifts to the newer paradigm: express unique originator co=
ntrol over the flow of attributes to each potential using site.

2. if you use the XRI form of the identity URI (rather than a more classica=
l HTTP URL), a service like http://freexri.com/ will act like DNS and resol=
ve your xri URL either into a more familiar but hard to type URL (@freeid*l=
ockbox resolves via forwarding to http://rapattoni.trustearer.com/lockbox).=
 Other resolver services will map one or more of your aliases into a blog-s=
tyle canonical id, registered for permaneny of reference. A standardized se=
rvice is the contact service - invoked like all other xri services uing wha=
t is essentially a crufty query syntax built into the URI name form. The en=
d result is like a FOAF file in concept, but has standard endpoints for all=
owing you to maintain n parallel friend-of-friend lists, where the process =
of update is standardized. I still investigating practice here, to see if a=
nyone has used RDF or semantic markup in their rendition of this service.  =
The nearest Ive come to seeing the combindtion of XRI + openid + semweb is =
at http://www.ootao.com/Page/XDIDemo.html where a broken demo once delivere=
d XRI  + LID (where Im guessing that the ref to  LID is to one of the fore-=
runner of the openid effort). A another partially broken demo at freexri.co=
m does showcase the XRI contact "service"; once they fix the openid asserti=
on making capability, it may be a close rendition of the ootao prototype.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.usefulinc.com/pipermail/foaf-dev/attachments/20080325/274=
23985/attachment-0001.htm