Re: tone of discussion regarding XForms/WF2

On Mon, 4 Sep 2006, Mark Baker wrote:

> On 9/4/06, Henri Sivonen <hsivonen@iki.fi> wrote:
>> So to rephrase the question:
>> How could the unification be accomplished without defeating a
>> significant part of the essence and value proposition of WF2?
>
> That's my concern too.  I suppose now would be a good time to call 
> for proposals to find out; post 'em if you got 'em, as they say.
>
> Mark.

Hi Mark,

We need an exposition of what the essence and value proposition is 
for WF2.  My assumption is that WF2 is just a proposal and not 
something cast rigidly in stone. We need understand what developers 
are looking for in simple to complex forms and for novices to 
experts. Here's my cut at what WF2 offers:

   - incremental extensions to HTML4 syntax
   - autocompletion support
   - text wrapping control
   - small set of core data types
   - some simple constraints (e.g. min/max)
   - calculated values (output control)
   - repeating controls
   - flat data model
   - use of scripting for validation
   - fields can belong to multiple forms

WF2 doesn't appear to address the need for structured data as is the 
case for both XForms and VoiceXML. WF2 also embodies the idea that 
fields own the data, and hence can appear as part of multiple forms. 
In XForms, fields are just views onto the data, using the well 
established model-view-controller design pattern to separate the UI 
controls from the data.

It seems to me that if we want a forms-lite then it should be
a well defined stepping stone to the conceptual structures
needed for more complex forms. In my view, a forms-lite 
proposal should address things like:

   - simple syntax
   - structured data
   - small set of data types
   - integrity constraints
   - calculated values
   - autocompletion and wrapping
   - ability to select which data is submitted and how
   - labels, hints, help and alerts

It shouldn't be too hard to come up with something that is 
essentially syntactic sugar for a subset of XForms. This will 
probably involve choosing a syntax where data types and simple 
integrity constraints are represented as attributes on input 
controls, where the data model is implicit. The XPath subset
used in XForms seems simple enough for newcomers to learn without 
difficulties. The end result would not be WF2 as it currently 
stands, but would address the needs for some syntactic sugar for
the full blown XForms syntax and build upon WF2, XForms and
VoiceXML, all of which have valuable ideas to contribute.

I will provide more details in my next email.

  Dave Raggett <dsr@w3.org>  W3C lead for multimodal interaction
  http://www.w3.org/People/Raggett +44 1225 866240 (or 867351)

Received on Tuesday, 5 September 2006 10:08:39 UTC