W3C home > Mailing lists > Public > www-xsl-fo@w3.org > February 2001

Re: extensions to FO

From: David Carlisle <davidc@nag.co.uk>
Date: Wed, 7 Feb 2001 11:11:09 GMT
Message-Id: <200102071111.LAA22036@nag.co.uk>
To: mf@w3.org
CC: www-xsl-fo@w3.org

> Raise an error and continue, or die?

I think I meant die, although as I say it really is up to the spec
for a given language to say what's supposed to happen.
If using schema you can., as I recall choose to allow or disallow
stuff from other namespaces at any given point.

>  As for the spec explicitely
> allowing foreign namespaces I don't see how this could work

Isn't this what 
http://www.w3.org/TR/xmlschema-1/#namespace_constraint
gives you options for doing or not doing?

> Now in the case of XHTML, XSL, MathML and others, listing what to do
> with other (explicitely listed) namespaces should certainly not be in
> the spec 

XHTML is very explicit that other namespaces are not allowed (unless you
have built a DTD, as I have for XHTML+MathML that allows a given element
(math in my case) in the content model of specified XHTML elements)

> I agree with Sebastian
> saying that the specs should only tell us the syntax and semantics of
> the FO namespace.

I agree that it is useful for FO (and probably Fo spec should say
something about it) and that's what I implemented in xmltex, but
I don't see anything that suggests any global rule about how a processor
for one namespace should treet elements in another. Die seems to be a
valid response. xmltex ignores the element and processes content
(which is what an XHTML browser will most likely do as well)
You say you's expect that the element and all its children should be
skipped. That would actually be quite hard for xmltex. Skipping
character data is not what its optimised for, it really wants to typeset
it:-)


David
Received on Wednesday, 7 February 2001 06:12:15 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 3 October 2007 16:06:08 GMT