[Dev] Some quick comments on DataModelSchemaXML

[Micah Dubinko]

This is about the format described at
http://wiki.osafoundation.org/bin/view/Main/DataModelSchemaXML

The following are some quick comments on this format, having worked
with it for an hour or two.

Overall: this is an excellent format. I could easily work with it (and
will be soon, with Noteliner)

Case: I would probably find myself forgetting which elements start with
Upper Case vs. lower case. A system that's worked well for me is:
* all elements user lowerCamelCase, but try to come up with such short,
one-word names that elements are effectively all lowercase
* attributes are all lower

id attributes: There might be some benefit to making these "officially"
of type ID, though it's currently hard to to that without directly
connecting a DTD or XML Schema. The colon character is allowed in ID
values, but it doesn't have any special meaning (which could be
confusing, looking at the examples)  Requirements for "xml:id" are
currently at Working Draft. http://www.w3.org/TR/xml-id-req/

If the source document really makes use of itemref/uuid
interchangeably, the XSLT I posted earlier would get alot trickier. If
we don't need the flexibility, I'd say pick one way and stick with it.

Inclusion: I'm generally against piling on things that require xmlns
declarations at the top, though xml:include might be a standard way
(already supported in some XSLT engines) to handle inclusions.
http://www.w3.org/TR/xinclude/

Taxon: I could live with that term.

AttributeDefinition: I don't have any major problems with it as
defined. A shorter, one-word name (if possible) would be nice. :)

Thanks,

.micah

=====
_XForms Essentials_ now in-print and online at
http://dubinko.info/writing/xforms/

__________________________________
Do you Yahoo!?
Yahoo! SiteBuilder - Free, easy-to-use web site design software
http://sitebuilder.yahoo.com