This is an archived snapshot of W3C's public bugzilla bug tracker, decommissioned in April 2019. Please see the home page for more details.
I think that you need to say something about ambiguous RELAX NG schema that could lead to conflicting rules, such as for instance: <choice> <element name="p"> <its:schemaRule its:translate="yes"/> ... </element> <element name="p"> <its:schemaRule its:translate="no"/> ... </element> </choice> Of course, this is a simplistic example and in practice, ambiguous schemas can be much more complex to detect and rewrite. Eric
(In reply to comment #0) > I think that you need to say something about ambiguous RELAX NG schema that > could lead to conflicting rules, such as for instance: > > <choice> > <element name="p"> > <its:schemaRule its:translate="yes"/> ... > </element> > <element name="p"> > <its:schemaRule its:translate="no"/> ... > </element> > </choice> > > Of course, this is a simplistic example and in practice, ambiguous schemas can > be much more complex to detect and rewrite. > > Eric yes, we need to do say s.t. about that. Actually we have not addressed processing of schemaRule elements at all so far (e.g. how about <p> declared as part of <text> versus <p> declared as part of <footnote>? Who is responsible for finding out which <p> is meant? Do we mandate a schema processor for this? This question arises even without ambiguity. I don't know an answer yet, though.
We gave up the schemaRule element, so this bug can be closed.
Closed, no further action necessary.