Re: spec-prod, xmlspec, docbook and Co. (esp. metadata)

Norman Walsh wrote:
> 
> / Philippe Le Hegaret <plh@w3.org> was heard to say:
> | OASIS is leading towards the XML version of DocBook. The W3C started to use
> | xmlspec in 1996/1997 with the XML specification itself. I didn't look
> | closely at DocBook and I'm still wondering how far we go into producing
> | a common schema for specifications, but is there any chance that we can merge
> | both somehow (or use a correct extension mechanism)?
> 
> Considering <article> in DocBook to be roughly equivalent to <spec>,
> I arrive at the following comparison (off the top of my head):
> 
> DocBook Articles vs. XMLSpec
> 
> - The "meta" is quite different, XMLSpec has a whole bunch of W3C-specific
>   metadata. This could (should?) be addressed by creating an XML namespace
>   for the W3C metadata. The DocBook TC is evaluating what to do with meta
>   and allowing namespaced meta seems like a good idea.

Regarding metadata, I've done a bunch of work in this area.
I scraped our TR index into RDF:

  http://www.w3.org/2000/04/mem-news/tr2.rdf
technical details on how it works are described, at least
in a way that the machine understands, in...
  http://www.w3.org/2000/04/mem-news/Makefile
I haven't written it up for people ;-)

The format, in brief, is:

    <REC rdf:about="http://www.w3.org/TR/1999/REC-xpath-19991116">
        <dc:date>1999-11-16</dc:date>
        <dc:title>XML Path Language (XPath) Version 1.0</dc:title>
        <doc:versionOf rdf:resource="http://www.w3.org/TR/xpath"/>
        <editor rdf:parseType="Resource">
            <contact:fullName>James Clark</contact:fullName>
        </editor>
        <editor rdf:parseType="Resource">
            <contact:fullName>Steve DeRose</contact:fullName>
        </editor>
    </REC>

using a whole bunch of namespace that we're working
on in Semantic Web Advanced Development:

<rdf:RDF xmlns="http://www.w3.org/2001/02pd/rec54#"
    xmlns:contact="http://www.w3.org/2000/10/swap/pim/contact#"
    xmlns:dc="http://purl.org/dc/elements/1.1/"
    xmlns:doc="http://www.w3.org/2000/10/swap/pim/doc#"
    xmlns:h="http://www.w3.org/1999/xhtml"
    xmlns:log="http://www.w3.org/2000/10/swap/log#"
    xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
   
xmlns:v="file:/home/connolly/w3ccvs/WWW/2000/04/mem-news/tr-merge.n3#">

The rec54 namespace is some work I've done on codifying
the W3C process in RDF:

  http://www.w3.org/2001/02pd/rec54
  http://www.w3.org/2001/02pd/rec54.n3 -- in RDF/n3
  http://www.w3.org/2001/02pd/rec54.rdf -- in RDF/xml
  http://www.w3.org/2001/02pd/rec54.png -- diagram in PNG
  http://www.w3.org/2001/02pd/rec54.svg -- diagram in SVG

The diagram tools are something I've written up for people:
  http://www.w3.org/2001/02pd/
(Norm, that's the stuff I showed you in Cambridge).

I've also got a script to convert the IETF RFC index
to RDF,
  http://www.w3.org/2001/02pd/rfcIndexGrok.pl
and a schema/model of their process:
  http://www.w3.org/2001/02pd/rfc65
  http://www.w3.org/2001/02pd/rfc65.n3

This stuff started as a result of automating the list
of URI schemes...
  http://www.w3.org/Addressing/schemes
I started thinking about automating the maintenance
of it... new URI schemes come from RFCs, so I needed
an automated source of RFC metadata... and of course
it seemed silly to have the IETF's digital library
described in RDF without doing W3C's own, so I did that;
I'm hoping it's useful to demonstrate stuff to
the dublin core community... the IETF has its metadata,
W3C has ours, and they're related (by, e.g. rdf:subPropertyOf
assertions) in a way that we can exploit in RDF
query engines. Danbri, did your work on this cross-vocabulary
query stuff come together?

Meanwhile...

Dominique Hazaël-Massieux, our webmaster, has been working
on a tool to chec W3C publications w.r.t. our publication
rules...

  http://www.w3.org/2001/07/pubrules-form
  http://www.w3.org/2001/07/pubrules-checker.xml

We're working on enhancing it to produce metadata
in RDF ala the above as a byproduct of checking.

So whatever format folks use to write specs, pls
consider the RDF format above for metadata, and/or
consider an XSLT gizmo to extract RDF metadata,
and/or just before you can generate XHTML
per W3C publication rules, so we can get
the metadata out that way.

[... more on other stuff separately...]

-- 
Dan Connolly, W3C http://www.w3.org/People/Connolly/

Received on Wednesday, 17 October 2001 11:31:06 UTC