RE: IETF input to XML Schema 1.1

["Larry Masinter" <LMM@xxxxxxx>]

> I'm concluding that I've asked "the IETF" as best I can and the reply is 
> "Thanks for asking (you are after all a polite bunch:-), but XSD didn't do

> it for us and we're now happy with Relax-NG."

> Not necessarily the desired answer, but none the less a good clear, 
> (actionable) answer.

I don't think that's the best conclusion to draw. Rather, I think
it's worth actually figuring out what kinds of extensibility people
needed that they got from RelaxNG, and you'll likely find different
answers for different applications. There are also some groups who
used XSD 1.0, so I don't think "IETF" has endorsed RelaxNG -- not
at all. At one point in the drafting of the guidelines, we had
a recommendation to use XSD and were "encouraged" to remove it,
because it was clear that the topic was controversial, but we
didn't recommend RelaxNG either.

It seems that many working groups have debated the XSD vs. RelaxNG
question in their context, it's worth going and reading those public
archives to tease out what kinds of extensibility they actually
needed.

I just think it requires some research.

My advice to the XML schema working group is to set a higher goal:
for all practical applications, be as extensible, as clear, and 
as simple to write and understand as RelaxNG.

If XSD 1.2 is just a step along the path, at least you'll have
a roadmap.

Larry