[rdfweb-dev] new FOAF spec -- preview

Dan Brickley danbri at w3.org
Mon Aug 11 23:54:27 UTC 2003


OK, caveats first.

This is rough (you're used to that though), the lead-in prose needs reorg 
and examples, the CSS might not work for everyone, and the table of
contents etc aren't there. The URL will 404 eventually, too. I am a bit
wary of this doing the Weblog rounds since a cleaned up version would be 
so much worthier of people's attention, but delaying further was worse.

In the tradition of release early and often (anyways some folk
from #foaf IRC have already seen it) here is a work-in-progress towards a new 
and less sketchy spec. I'm also snowed under the rest of this week so 
won't have any evenings to hack on it. (Hopefully next weekend...)

http://xmlns.com/foaf/0.1/_tmp

As you'll see, the bulk of the document is what I've misleadingly called 
'Appendix A', a cross-reference of FOAF terms, with domain and range for 
each property, and back-pointers from properties to classes. The doc
itself is assembled by combining information from the RDFS/OWL FOAF 
definitions, http://xmlns.com/foaf/0.1/index.rdf with the one-per-term 
HTML fragments in http://xmlns.com/foaf/doc/ -- note that these are
named things like 'knows.en' in the hope that someday we'll have
translations of these into other languages.

There is a machine-generated 'FOAF terms by A-Z' listing, and a
hand-maintained 'FOAF at a glance' alternate view, which just about 
squeezes it all into a page.

So the cross-referenced term descriptions are the main thing we need to 
discuss, at this stage. There is one for each class/property currently 
described in index.rdf and the build scripts (see [1]) will gripe if 
I try to commit a copy of the spec that has an undocumented term. It
should soon be reasonable to expect the FOAF spec to describe every 
FOAF class and property. Well, it's reasonable already, we just don't 
meet that expectation.

Now that we have the vocab status annotations, 
http://www.w3.org/2003/06/sw-vocab-status/ns# (eg.
vs:term_status="unstable") in the schema, we should also be able to be 
less wary about documenting the more experimental bits of FOAF, since
they can be tagged with a health warning. There should also be a 'by
status' section of the spec, as well as a full account of what we mean
by 'stable','testing', 'unstable'.

Anyway, some progress; more needed. Feedback as ever welcomed via
rdfweb-dev at vapours.rdfweb.org.

cheers,

Dan

[1] http://xmlns.com/foaf/0.1/specgen.rb
    --script used to generate cross-reference section

ps. thanks to Cardinal and Nicole for CSS advice :)    



More information about the foaf-dev mailing list