[rdfweb-dev] fwd: Broadband Mechanics requests and proposals

[Dan Brickley]

Fwd'ing this from Marc; it went to the list -bounces address by
accident.

Lots of questions that deserve answers, too many to attempt to answer
them tonight.

I've just lost a week of dayjob time + evening + weekend. Hopefully I'll 
be able to crank out a reply to most of this tommorrow. 

One quick point re versioning: 'http://xmlns.com/foaf/0.1/' just is the
FOAF namespace URI. Try to ignore the '0.1' or think of it as a bad joke 
(1.0 backwards). There will not be changes to the namespace URI simply
to celebrate the alleged stability of (bits of) the vocabulary. This was
a clear lesson from the Dublin Core Metadata Initiative experience with 
versioning. In DC, we had things more or less right and were using 
http://purl.org/dc/elements/1.0/ then 'cos of a few modest changes to
the definitions talked ourselves into trying to upgrade to 
http://purl.org/dc/elements/1.1/ 

That sort of migration proves to be a practical nightmare. The approach
in FOAF instead is to think of the namespace as a slowly growing
dictionary where terms can be looked up, and term definitions rise
through maturity levels of 'unstable', 'stable', 'testing'; but do so
without their identifiers changing.

I should also address in more length the question of 'who is in charge 
around here' w.r.t. to the core FOAF vocabulary. The short answer is
that it is myself and Libby Miller. The longer answer doesn't let the
rest of you off the hook so easily... And the even longer answer would 
of course mention RDF's inhererent extensibility empowers anyone to
create extensions that can be mixed with core FOAF.


Relatedly, I have a half-started 'Fear of a FOAF planet' piece in progress for the
weblog, on questions of adoption, uptake etc., and whether to completely 
remove the wilful obscurity that has given us a quiet place to work
these last couple of years. There are a number of things I want to get
right before really encouraging mass adoption of FOAF. This list belongs 
in Bugzilla, but includes (i) privacy policy and guidelines for users
and aggregators (ii) accountability and best practice for aggregators, 
including bot behaviour and presentation of personal data in public
sites (iii) vocab stability esp w.r.t. I18N -- finishing of the foaf
naming discussion. Until we have at least nailed, plus a few other
things, I feel we would be irresponsible to go looking for real mass uptake,
Slashdotting etc. All that said, that day draws inevitably closer, hence
various efforts to get things into shape... 


Bedtime thought ;)
  
   If we were slashdotted tommorrow, how many FOAF sites would collapse?

Dan
-------------- next part --------------
An embedded message was scrubbed...
From: mailman-bounces at vapours.rdfweb.org
Subject: Uncaught bounce notification
Date: Tue, 22 Jul 2003 01:03:22 +0100
Size: 37212
Url: http://lists.usefulinc.com/pipermail/foaf-dev/attachments/20030721/c05ae1d7/attachment-0001.eml