[
Lists Home |
Date Index |
Thread Index
]
> Weren't such corner cases caused by a lack of formal
> underpinning and complexity of XSD spec?
Complexity, yes. Lack of formal underpinning, no. I don't think one can
argue that Relax-NG has stronger formal underpinning than XSD has.
For the rule in question, there is no ambiguity in the specification. In
fact the rule is very simple. It's just very easy to miss it, given the
astonishing complexity (one might say, formality) of the overall rules for
handling referential integrity constraints (which, as far as I can tell,
have no parallel in RELAX-NG).
I'm surprised though that Xerces doesn't catch violations of this particular
rule: it's normally very conformant.
Michael Kay
http://www.saxonica.com/
|
