Re: Fw: XForms requirements

At 04:52 PM 26/03/2001 -0800, John J. Barton wrote:
>Discussions of XFORMs and the documents about it
>take a rather narrow XFORMs-centered view, but I don't
>think that the technology is exclusive.  Rather, I think
>XFORMs glues together data-input (downloaded instance)
>data-output (form submission), and data-update (UI+Logic)
>elements that can be mixed with other elements. Dialog
>boxes and multiple views are not defined by XFORMs but
>neither are they excluded.

True, but allowing non-standard extensions is not the same as providing a 
standard that all application programmers can rely on all clients 
supporting. Nor does it seem sensible to me that in order to define a UI, I 
have to put part of it into XUL and the other part into XForms.

XForms has a really nice architecture, and it is 90% of the way towards 
providing a rich enough user interface that it can be used for 99% of the 
applications out there (remembering that 97.463% of all statistics are made 
up on the spot). I'd just like to see it make that last push.

>  Their absence in the XFORMs
>spec. is a result of componentization.

Does this mean that these added UI elements would be supported in XForms 
Level 2? Or that there is another UI definition working group complementary 
to XForms that is extending its work?

I've looked at UIML and XUL, and neither one seems to do the job. UIML is 
way too abstract, and XUL looks far too specific to a heavy client 
environment without the XForms-like features one would expect.

What I really want is the ability to define UI's as data, and to have a 
generic client that can run any application just by pointing it at a UI 
configuration file, preferably a client lightweight enough to be supplied 
as a download-on-demand applet. Defining the UI as data is not a new idea, 
it has been around at least since FoxPro introduced its Forms Designer in 
the late '80s, but XForms has the potential to make it standard and 
available to everyone in a way that promotes View separation.

Received on Tuesday, 27 March 2001 14:10:41 UTC