> I think the prominence of RNG and Schematron examples in the
> specification mitigates this. Those whose assumptions are
> driven merely by the syntax of one PI are not even, in my
> estimation, going to be able to understand W3C XML Schema
> enough to use it.
You're not likely to have to deal with bug reports that say "I wrote
<?xml-schema books.xsd?> and Saxon ignored it."
But I am. I'm concerned about the people who understand very little, and
there are lots of them. These people see an example on a web page and copy
it, guessing the semantics from the keywords used. We really need to avoid
creating such obvious usability traps: the XML space has far too many of
them already.
Regards,
Michael Kay
http://www.saxonica.com/
http://twitter.com/michaelhkay
