Re: Irony heaped on irony

At 05:04 AM 5/19/00 -0500, Dan Connolly wrote:
>"Simon St.Laurent" wrote:
>> I'm deeply concerned that this rush toward the 'self-describing web' is
>> taking us closer and closer to the 'self-destructing web', as a large
>> number of controversial issues are being glossed over in the interest of
>> making namespaces point to schemas, today.
>
>Hmm... I have a hard time seeing this a rush. 

The main aspect of it that I see as rushing is this quiet shift to using
namespace URIs to point to schemas.  This controversial practice is being
put forward through the schemas group, for which its useful, with no
apparent discussion of its impact on other XML activities inside and
outside the W3C.

This activity is a large part of what makes the relative URI discussion so
ugly, because the results of that are intertwined with the
NSURI-pointing-to-schemas issue.

>The ideas have been
>around and discussed in the abstract here and there for years
>and years. 

As Larry Masinter said earlier, "In theory, theory and practice are the
same.  In practice, they're not."

>We made a big step with XML 1.0 where the element structure
>of a document is clear whether you look at the DTD or not,
>which allowed us to consider using something other than DTD syntax
>to describe the structure of XML documents, which allowed us
>to start work on XML Schemas.

It also allows people to chuck schemas altogether or use schema models
developed outside the W3C, something that the W3C doesn't seem so keen to
encourage.

>We spent a year talking about the details of using the Web to
>find out about markup vocabularies during the development of RDF
>and the Namespaces spec, and another 12+ months during the development
>of XML schemas.

Who exactly is this 'we'?  Discussions on XML-Dev haven't taken this turn,
and the overlapping W3C/XML-Dev folks don't seem very fond of it.  If the
discussion is taking place inside the W3C only, I'd suggest that it might
be easier to build a case for such infrastructure by working in public.

>Now we have an XML Schema spec that is (hopefully) complete,
>if not entirely clear yet, and we have a bunch of coding going on.
>If the coding and stuff shows that it won't work, I'll be
>disappointed, but I'll have to accept that. On the other hand,
>if the coding and stuff shows that it does work and it's useful,
>then that will speak for itself too.

I'm not sure that this justifies the use of Namespace URIs to point to
schemas.  Schemas may be useful (or not) without that piece.

>> The particular path being proposed to the 'self-describing web' is of
>> questionable value, built on a spec that's been controversial since its
>> first appearance and which continues to raise new questions.
>
>I have no questions about the value of a self-describing web. I accept
>that you have questions, and that you're not alone. What better way
>to address these questions than to try it out?

I'd suggest that we try it out in ways that don't affect development across
the Web as it currently exists, in a smaller (but open) laboratory
environment which is kept separate from the specs people use to get work
done today.

I'd love to try it out - but let's try it out in ways that don't block
other possibilities, and which cause least harm if there are giant hiccups
along the way.

>> As much as people hate it, maybe it's time to slow down, get some
>> experience with how these things work in practice, and move toward the
>> 'self-describing web' when (and only when) the parts involved are
>> well-understood and considered clean.
>
>Well... here we are... just about done with XML Schema last call
>and getting ready for Candidate Recommendation, exactly the
>part of the W3C process intended to gather implementation
>experience. If the experience exposes problems, then we'll
>have to think some more.

I think the experience has already exposed problems, as the W3C has started
putting schemas at the locations referenced by their namespace URIs and
received some loud comments from people that this may not be so good an
idea.  Is this experience part of testing XML Schemas?  I'd argue that it
shouldn't be, that it in fact reflects a change in practice more relevant
to namespaces than schemas.

>But I hope you aren't suggesting that we all just think about this
>stuff in the abstract until we get warm fuzzies. I intend
>to code up some stuff and work out the details in practice.

Sounds like it's too late.  Develop in private, work out details in public
practice, and never mind what the side effects on non-W3C Web practice may be.

Not a good recipe, IMHO.
 
Simon St.Laurent
XML Elements of Style / XML: A Primer, 2nd Ed.
Building XML Applications
Inside XML DTDs: Scientific and Technical
Cookies / Sharing Bandwidth
http://www.simonstl.com

Received on Friday, 19 May 2000 11:28:43 UTC