> At this point, I'd say no to internal XSchemas, if only because sorting out
> those details will take a long time and considerable debate. It would also
> require us to find some way of hiding the internal XSchemas from non-XSchema
> aware processors, something I'm not fond of.
I agree with Simon. I can't think of anything you can do with an internal
XSchema that you can't do with an external XSchema, so including them in 1.0
just means having to wrangle over a bunch of rules for resolving duplicate
XSchemas. With the recent partial validation discussions, this gets even
uglier...
-- Ron Bourret