Re: Problem in mathml2 schema layout.xsd

> However, the way notation is defined in layout.xsd, only one value is
> allowed.

Thanks for this fix.

> However, would we run into problems when applications using the published schema take
> our xml? What would you suggest we do?

There may be some applications that don't render menclosed (even if you
only use one term) but the spec is unambiguous here and the schema
(which is not normative) is clearly more restrictive, so you should
validate to your locally fixed schema and we should consider updating
the one on w3c at some point or (depending what the group decides) at
least make sure it is fixed in the mathml3 schema.


Actually like many MathML attributes the list is defined to be open.


> 3.3.9.2 Attributes 
> ...
> The values allowed for notation are open-ended. 

This is why the DTD (which is normative) just types it as

      notation CDATA 'longdiv' >


ie (in xsd terms) just xs:string.

Howver while an open schema is good for implementers to experiment with
interesting extensions it's not so good for driving schema driven
editors and other systems into helping the author avoid or detect
errors, so the schema is (intentionally) over restrictive in some places
however in this case I would say that your version, allowing a white
space list from a fixed set is probably the right level of restriction
in this case, or maybe just relax it to a list of tokens, depending on
the use case here.

David

________________________________________________________________________
The Numerical Algorithms Group Ltd is a company registered in England
and Wales with company number 1249803. The registered office is:
Wilkinson House, Jordan Hill Road, Oxford OX2 8DR, United Kingdom.

This e-mail has been scanned for all viruses by Star. The service is
powered by MessageLabs. 
________________________________________________________________________

Received on Wednesday, 16 July 2008 08:12:31 UTC