RE: xforms

Hi David,

Thanks for your great comments:

>Terminology: Why facets?

That terminology is from XML Schema. Even though it was poorly specified in
the first draft, facets are not mere attributes. A definition section would
be a worthwhile addition to the document, no? :-)


><quote> It is recommended that user agents offer date and time pickers...
>None of those I have seen have been very accessible?

That may well be. This will be dealt with when we get to the User Interface
document. The data model provides a solid foundation for building accessible
UI on top of.


><quote>Binary data could be packaged either in-place as part of XML form
>data </quote>
>What of the charset defn of XML? 

That perhaps could be clearer, but "packaged" binary data would be base64,
etc..


>5.3 Could lead to horrible complexities...
><string name="spouse" required="status is 'married'"/>

Yeah. Some of our best men are working on improving that one. :-)


>5.4 As above, question robustness of cross references.

Again, the expression language is in a particularly early stage. We are
investigating other approaches (like XPath) and would welcome specific
comments. As you noted, one thing we do allow is an escape to a local
scripting language.


>5.5 Will 'form filling' not be a valid machine to machine operation?

We may be using different terminology, but according to our requirements
document, saving and resuming will be a part of XForms. How validations
apply to a machine to machine transaction hasn't yet been spelled out.


>6.3 <q>Groups ... the syntax: customer.street. </q>
>Should a data model make such  blatant implementation oritented statements?

It doesn't really bother me. One part of our work is to define a forms DOM,
so referring to it from another section seems OK.


>[duplicating groups]
>Isn't the entity usage ideal for this? 

We're always open to possibilities, but I suspect that entities wouldn't
meet our needs, the DTD subset angle in particular.


>7.x I'm curious why you have adopted an Ecmascript approach

I'm not sure what you mean by "an Ecmascript approach". All I can say is
that this is currently one of the least developed parts of the document.
Much work remains. Our next release of the Data Model Working Draft should
answer most or all of your questions.

Thanks,

Micah Dubinko
Editor, XForms series of documents

Received on Monday, 1 May 2000 19:38:45 UTC