[rdfweb-dev] foaf as semantic web sampler

[Bill Kearney]

> I've been thinking about Bill's comment re schema bloat, and why it
> seems to me to be ok to have a few things in FOAF like foaf:myersBriggs,
> while not trying to be completist in our coverage of (say) personality
> tests or location/address info.

If the examples, specs, website and/or wiki make statements to this effect it
will help. It'd also possibly harm things. But to list it as being in a state
of flux and open to discussion and/or evolution WILL help. Better to let folks
know it's a work in progress /that's usable/ but a step toward new things.

> It boils down to the notion that FOAF is a sampler, or advance-warning
> teaser, for the Semantic Web. We do enough to give people a flavour of how
> things might be, and enough to flush out some of the technical and social
> issues raised by RDF/SemWeb. Pursuing this, and trying to build something
> fun at the same time, it is quite reasonable to have the mix of classes
> and properties we seem to be evolving for FOAF, while not trying to be
> completist.

Right, it's in that complete-ist attitude that people get alienated. Thus
extensibility being valuable. If I can, from the start, graft on "extra stuff"
that's cool to me, I might be able to get others to buy in the idea. This is
the whole extensiblity argument so I'm just restating the obvious. The trick is
balancing between just grafting on new namespaces, trying to add to existing
namespaces or having a fluid core schema that's not formally tracked anywhere.
We have all three right now as RSS has evolved. My interest in /not/ putting
MBTI directly into foaf is to motivate a means to express extensible metadata
without requiring new element namespaces.

There's a set of mental stumbling blocks that often impair new user adoption of
all this. The most confusion (imnsho) being namespace URI does not mean URL
argument. Followed quite quickly by the RDF you can't have contents inside and
element distinction. The combination of these two continues to confound me and
I'm sure others get headaches from it as well. So my thinking is to hijack some
of the semantics from efforts like HTML meta tags to give new users a way to
"grok" the purpose of this stuff. That is without having them wander down the
road of inventing entirely new namespaces that reinvent other work (often
badly).

Anyway, it's cool stuff and it's great to see the open discussions continuing.

-Bill Kearney

> BTW there's a _big_ list of tests at http://www.intp.org/tests.html that
> you might find fun...
>
> Righto, back to fiddling with the hard drive booting setup for my PC. There
> goes the weekend...
>
> Dan
>
> To unsubscribe from this group, send an email to:
> rdfweb-dev-unsubscribe at egroups.com
>
>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>