Re: IBM Position Statement on XForms and Web Forms 2.0

Hi Raymond and Francisco,

Please let me start by saying that nothing should be read into the timing 
of the release of the IBM position statement.  The statement was 
formulated over a many week period that included Ian Hickson in a lengthy 
thread of discussion on the W3C chairs list.  The formulation of the 
statement also included consultations with Steve Bratt and Chris Lilley, 
and it obviously included internal consultations within IBM.  We had no 
knowledge of Ian's pending vacation.  However, a few choice quotes from 
Ian on the chairs thread are:

"Web Forms 2.0 is an XML language" (August 18, 2006)

"The first problem with shoe-horning XForms into HTML is that XForms is 
based on XML, and HTML is not. We can't require an XML-based language" 
(August 17, 2006)

"I was not involved in the decision to take this [Web Forms 2.0] to the 
W3C; indeed I unsuccessfully attempted to discourage it when I heard about 
it." (August 18, 2006)

"the WAF working group came to me and asked if they could publish it and 
whether I would be the editor if they did so, a task that I was happy to 
take on; I think Opera, one of the group's members, was the main proponent
" (August 17, 2006)

There were similar contradictions on the technical side of that chairs 
thread, such as claiming that Web Forms 2.0 works in IE with a plugin but 
that XForms doesn't (but not giving XForms the benefit of a plugin).  Or 
that XForms would be OK if it were *only* allowed to have input, button 
and select tags, while simultaneously allowing WF2 to have whatever other 
tags it needs to duplicate the functionality in XForms.

Another mystifying claim to me is Raymond's assertion that we would break 
backwards compatibility of XForms document by *adding* features that 
unified the best of what Web Forms 2.0 has to offer into XForms.  How did 
you come to that conclusion?  *Backwards* compatibility means that old 
document continue to work in new processors, not that new documents 
created with a *higher* version of the language will work in old 
processors?!?

This is the kind of stuff that we need, as a W3C team, to stop doing.  WF2 
has some features in it that we have been considering for the future of 
XForms anyway because we really do want to smooth the migration path for 
content.  WF2 also contains a lot of quite unnecessary divergences of 
concept from XForms.  We need to get rid of those.  We really can't afford 
to bifurcate HTML in ways that can never be made compatible.  We also need 
to pull together and reexamine, without hyperbole, whether the team can 
come up with a way to preserve the XML basis that has become the 
foundation of HTML over the past six years.

It will be necessary for all involved to not cling so tightly to positions 
that there is no room for the compromise that's needed.  Remember, 
perfection is the devil of excellence, and since we're already talking 
about how various UAs will gracefully degrade on various kinds of content, 
it seems viable to move all parties to a common ground on what constitutes 
being "graceful enough".  This will involve focusing on the *main* use 
cases as actually being more important than *edge* cases.

In conclusion, you asked what you are supposed to make of the IBM position 
statement.  Gosh, I really don't know why it is 'incomprehensible'.  It 
seems to say:

1) XForms has good features
2) WF2 has good features but significantly overlaps XForms
3) W3C team was supposed to require (in charter) the XForms and HTML 
groups to address the WF2 content
4) We strongly encourage the WF2 work to migrate to the XForms and HTML 
groups as prominent features of their new charters.  We want these groups 
to be compelled to solve the problem, and we want the interested members 
from WAF to participate in that solution.
5) We don't agree with the WF2 claim that XForms is for server-side and 
WF2 should be the on-the-wire standard.
6) XForms implementations have already demonstrated its viability 
on-the-wire.
7) XForms needs to do a better job going forward to profile for reduced 
functionality clients and the on-the-glass authoring model
8) Although easing implementation for browser vendors is one concern, the 
web is about a much broader spectrum of web technologies that are 
important to the whole consortium, so it is important to do our best to 
preserve the XML basis for new features to help entice content toward 
well-formedness.
9) There will be compromises needed from XForms and HTML as well as from 
WF2, but the benefits to the future of the web will be worth the work.

Please see my blog posting today for further context about this situation.
http://www-03.ibm.com/developerworks/blogs/page/JohnBoyer?entry=xforms_web_forms_2_0

And more importantly, let's get more comments and opinions, to'ing and 
fro'ing, and acrimony and sympathy until we find the common ground that is 
going to make the unification of XForms and WF2 a reality.

Thanks,
John M. Boyer, Ph.D.
Senior Product Architect/Research Scientist
Co-Chair, W3C XForms Working Group
Workplace, Portal and Collaboration Software
IBM Victoria Software Lab
E-Mail: boyerj@ca.ibm.com  http://www.ibm.com/software/

Blog: http://www.ibm.com/developerworks/blogs/page/JohnBoyer