[rdfweb-dev] FOAF as RDF (was Spring v1.3.1 and plans forFOAFspec improvements)

[James Carlyle]

Julian

> it's not meant to be abuse!). Now we can raise the water or lower the
> bridge. I've been pushing here on one end by trying to understand
> whether the standard can be brought closer to the XML world so that the
> many, many XML parser tools can be used. The big reason for this was my
> bad experience at the turn of the year trying to find a PHP parser that
> was stable and worked. At that time I failed. Anyway, it's pretty clear
> that you think bringing the standard to the tools won't work. Well OK,
> then FOAF adoption is going to be dependent on bringing the tools to the
> standard. Maybe I'm wrong but my perception is that there's still many
> programming environments that don't have a stable, mature RDF parser.

I see your point about how allowing a standard XML schema for FOAF would
open FOAF to many XML developers with XML tools.

But if FOAF could be expressed as an XML schema, then as surely as night
follows day, the RSS 0.9 --> 0.91 evolution would occur:

1) RSS 0.9 is described in terms of RDF, but specified as XML (Dan Libby or
whoever started with the RDF model, but the Netscape Netcenter spec is
described as an XML schema).
2) RSS 0.9 is implemented as XML by developers (i.e. the RDF model heritage
is lost).  For example, xmlTree's 1999 RSS parser worked at the XML level,
not the RDF level.
3) Time passes
4) The mass adopters of RSS argue that the RDF elements of RSS 0.9 are ugly
and unnecessary, and adopt plain XML RSS 0.91
5) The ability to mix in assertions becomes limited to whatever modules are
argued over and agreed (RSS 2.0)

I personally would like to see FOAF continued to be described in terms of
the RDF model, rather than a particular syntax.  I am curious to see what a
model-based specification will lead to, and whether we will end up with a
richer array of applications than we got for RSS (i.e. news aggregators and
blogging tools).  This is not to say that RSS tools are not worthwhile -
just that most seem to follow a limited pattern.  And RSS seems only ever to
be consumed in the context in which it was created.  For example, if RSS
content is created as news then it is normally consumed as news.  I am
curious to see if the Semantic Web can deliver something real, and if FOAF
assertions can be used in many different contexts.

Your point about not having stable RDF parsers for many programming
environments at the January FOAF meetup in London was what triggered me to
develop a global RDF XSLT stylesheet.  My logic was that almost all
programming environments can handle XSLT, even if they can't cope with RDF.
The stylesheet could be modified to "dumb down" RDF and convert the model to
a particular syntax if necessary, though at the moment it generates triples
and flattens the XML while preserving the model.

James