W3C home > Mailing lists > Public > public-appformats@w3.org > August 2006

IBM Position Statement on XForms and Web Forms 2.0

From: John Boyer <boyerj@ca.ibm.com>
Date: Tue, 29 Aug 2006 13:21:16 -0700
To: chairs@w3.org, w3c-ac-forum@w3.org, w3c-ac-members@w3.org
Cc: www-forms@w3.org, www-html@w3.org, public-appformats@w3.org
Message-ID: <OF28D35F01.C87771F8-ON882571D9.006B22EE-882571D9.006FD8FF@ca.ibm.com>
Dear W3C Chairs and W3C AC Members,

The text below represents the IBM position statement on XForms and Web 
Forms 2.0. 

An advance copy of this document was reviewed favorably by Steve Bratt and 
Chris Lilley, and I believe it is accurate to say that IBM offers this 
position statement in the spirit described by Steve when he said,  "I very 
much hope that it will serve to open (not close) doors to much improved 
cooperation from all parties to help us move forward.  I think this is a 
good start."

Based on further details provided by Steve and Chris on the rechartering 
plans in progress, I am left feeling quite optimistic that the full team 
of W3C working group members interested in this space will be able to 
forge a path of collaboration and compromise that satisfies the most 
urgent requirements of all the parties in a timely fashion.

Best regards,
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

===================================================================================

IBM is a strong supporter of the XForms recommendation and working group. 
As well, IBM has looked positively upon Web Forms 2.0 (WF2) since its 
member submission because the W3C team's acknowledgement strongly 
encouraged the charter mandates of XForms and HTML be enhanced to allow 
unification of WF2 features into XForms.

The current draft of WF2 suggests that we can appropriately meet users 
needs with two different technologies:  an XForms or similar system used 
purely at the server to meet the needs of powerful applications and their 
designers, and a different more HTML-oriented solution "on the wire".  On 
this point we respectfully disagree.   It is true that authors of simple 
documents and applications do indeed need a simple and convenient 
solution, but it is equally important that the technologies used in more 
powerful applications migrate seamlessly between client and server, 
according to the needs of the application and the capabilities of the 
client.  So the choice is not whether we can have XForms at the server and 
WF2 on the wire, but whether the "on the wire standard" can scale to meet 
the needs of both simple and powerful applications.  This is the reason 
that we are pushing so strongly for unification.

The first working draft of Web Forms 2.0 contains many good features, 
including XML Submissions, declarative repeating constructs, and strongly 
typed input fields, and indeed, many of the features exist in the XForms 
recommendation.  From this we conclude that many of these features are of 
use, not just at the server, but on the wire as well.  IBM sees a 
significant advantage for the future of the web to rationalize the WF2 
proposal with the XForms recommendation in the places where there are 
direct overlaps, while incorporating the WF2 features that are distinct, 
and being attentive to the needs of users who seek truly simple solutions.

XForms and WF2 have very many features and use cases in common, so it 
should be possible to unify the two with a common syntax.  Most of the 
current differences between XForms and WF2 appear to be along two 
orthogonal axes: 

1) Simple vs. Complex content: the extent to which new features of HTML 
can be accessed as simply as possible to scale up the functionality of 
content gradually. 

IBM believes this is a strength of Web Forms 2.0 that needs to be better 
incorporated into XForms.  For example, the XForms input control could 
accept the name attribute as a substitute for the ref attribute when 
creating the implicit data model, and attributes such as type, readonly 
and value placed directly on the input control can be used to implicitly 
create the appropriate model item properties and default values.  This 
type of change basically recognizes the legitimacy of the Web Forms 2.0 
requirements by importing the syntactic constructs directly with little or 
no modification.  Importantly, IBM understands that it is users (albeit in 
various roles, and with various levels of expertise) who must be 
reasonably satisfied with both the convenience and the power of whatever 
solutions we adopt.

2) XML vs. non-XML content:  the extent of belief in the premise that new 
features in HTML should be defined via XML syntax rather than non-XML 
syntax

While ease of browser implementation of new features is one important 
concern, IBM believes that there are many important factors that 
contribute not only to the total cost of ownership of web applications but 
also to the ability to create and maintain a significantly broader 
spectrum of current and future web technologies.  IBM believes that these 
factors are important enough that it is essential to use new features of 
HTML as an enticement toward greater conformance of web content to XML 
syntax, granting that effort will be required to achieve a result that is 
also sufficiently convenient for various users.

Simultaneously, IBM believes that the requirement of XML conformance for 
HTML that exercises new features does no real harm to legacy UAs, which 
would provide graceful degradation based on HTML content that happens to 
comply with XML syntax, and the requirement does no real harm to UAs 
updated to respond to the new syntax since they can be updated to respond 
to any new syntax.  As for accommodating reduced functionality UAs, we 
believe it is necessary to have a unified conceptual model from which a 
more appropriate language profile can be created for reduced functionality 
UAs.  The profile should focus on the ease-of-use syntax and its ability 
to imply a basic data model, and it should only delegate the more advanced 
features of an explicit model to the server when a reduced functionality 
UA is being served.

In conclusion, IBM strongly advocates for the renewed charter of the 
XForms and HTML working groups to include unification of the Web Forms 2.0 
work with emphases on the ease-of-use benefits from WF2 and the XML basis 
from XForms.  There will be compromises required of all parties, but also 
significant synergies that become possible by accommodating the full range 
of forms expertise available in the W3C.

===================================================================================
Received on Tuesday, 29 August 2006 20:21:55 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 8 January 2008 14:10:20 GMT