Re: HTML forms, XForms, Web Forms - which and how much?

It's pretty hard to understand how you (and Daniel) come to this 
conclusion given that the two working groups are *chartered* to come up 
with a hybrid approach, i.e. something that is NOT equal to XForms as it 
currently is written.

The WF2 crew have many times said "Since we already have a spec, why don't 
we start with that and you can point out what's wrong with it."  I would 
say the same thing.  XForms already is a W3C recommendation for forms. Why 
don't you start with that and point out what's wrong with it for your use 
cases?

I think if you hunt through all those use cases, you won't find as many 
things to fix as the hyperbole would suggest.

I think the biggest complaint is not wanting to use well-formed XML. Well, 
the new forms technology is supposed to cover that.  The XML serialization 
is called XForms transitional, but there should be a tag soup variety. 
It's the tags and their meaning that we want to be consistent regardless 
of minor lexical variations.

I think the second biggest complaint is wanting to reason "on the glass", 
i.e. wanting to let the UI controls suggest the data via their 'name' 
attribute.  This seems to be part of the plan too.  My main concern here 
is that it would be nice if UI hierarchy then implied hierarchy in the 
data, at least as an option, so that XML can come out of the end of the 
form fill process if XML data is desired. So, this is where we talk about 
including attributes like readonly, relevant, datatype and calculate 
directly on the controls.  It's also where we talk about what the repeat 
structure should look like.

Ah, now that I read ahead on the list, I see that Dave Raggett has asked 
essentially the same thing of everyone: let's talk more about design 
patterns and use cases [and less about sound bites].

John M. Boyer, Ph.D.
STSM: Lotus Forms Architect and Researcher
Chair, W3C Forms Working Group
Workplace, Portal and Collaboration Software
IBM Victoria Software Lab
E-Mail: boyerj@ca.ibm.com 

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





"Preston L. Bannister" <preston@bannister.us> 
Sent by: preston.bannister@gmail.com
04/30/2007 07:10 PM

To
John Boyer/CanWest/IBM@IBMCA
cc
"Simon Pieters" <zcorpan@gmail.com>, mark.birbeck@x-port.net, 
public-html@w3.org, public-html-request@w3.org
Subject
Re: HTML forms, XForms, Web Forms - which and how much?






On 4/30/07, John Boyer <boyerj@ca.ibm.com> wrote:

Hi Simon, 

SP: Sure, but when something can be done with either an existing concept 
or 
with a new concept, then using the existing concept seems preferable to 
me.

JB: Why would we ever write a language that allows one to say C = A + B; 
when we already have 

LOAD AX, 1000 
LOAD BX, 1004 
ADD AX, BX 
STO AX, 1008 


By the same token, why write C = A + B when you could write:

    (SETQ C (PLUS A B))

After all, Lisp is yet "more advanced" ... :)

John, I get the impression you think XForms is the best and only possible 
evolution and abstraction for generating HTML forms.  Seems there are a 
few folk (myself included) who are less convinced.  There may well be a 
range of applications for which XForms is a terrific solution.  There may 
also be a range of applications for which XForms is not the most efficient 
abstraction.