Re: XForms vs. Web Forms

On Wednesday 2005-03-16 10:26 -0800, John Boyer wrote:
> In essence, the WHAT-WG approach is a band-aid solution created by
> people who are not very familiar with the needs of forms applications.

The needs of forms applications in what context?  On the Web?  Or on
“Intranets”?

> Do they support the hijacking of the W3C process for bringing the 
> web to its full potential by supporting a rogue faction of
> *W3C members* who do not participate in the relevant working groups?

I think statements like this come from a misconception about what the
W3C is.  The W3C is a consortium of member companies.  There is nothing
in the W3C process to ensure that what the W3C does is designed
specifically for the Web (certainly nothing that does so effectively,
anyway).

If the W3C process ensured that things were designed for the Web, then:

 * There would need to be much less incompatibility between standards.
   (For example, SVG couldn't rewrite CSS, XForms couldn't be
   incompatible with the existing HTTP content negotiation regime
   without coming up with a reasonable alternative before reaching REC,
   etc.)  This requires producing fewer standards and producing them
   more slowly.

 * There would be more attention paid to how the Web would migrate to
   the new standards.  For example, XHTML 1.0 (first edition) would have
   defined when a user-agent should use XML rules [1] for parsing XML,
   and preferably in a way that would have allowed migration to XML
   within the text/html MIME type, which would have allowed the Web to
   switch to using XML for HTML.

 * Most importantly, working groups would have to take the candidate
   recommendation (CR) phase more seriously.  In particular, many more
   test cases would be needed, and for an implementation to count as
   interoperable for the purpose of exiting CR, it would have to
   correctly implement (potentially via plugins) all previous Web
   technologies (both correctly according to the test suites and in a
   way that is usable on the Web).  This would be needed to ensure that
   the various recommendations are compatible with each other.  (I'd
   claim that if the W3C worked this way, every W3C recommendation
   should still be in CR except for PNG, HTML 3.2, and XML 1.0.  But if
   these stricter CR requirements existed years ago, I'm sure many more
   would have met them by now.)

But the fact is, there's much more going on in the W3C than just the
Web, and those members that produce Web clients can't possibly
participate in all W3C working groups.  There are just too many of them.
And there's no way of telling which ones are relevant to the Web.

-David

[1] http://lists.w3.org/Archives/Public/www-html/2000Jul/0084
    http://lists.w3.org/Archives/Public/www-html/2000Sep/0024

-- 
L. David Baron                                <URL: http://dbaron.org/ >
          Technical Lead, Layout & CSS, The Mozilla Foundation