Re: Normalize / Simplify HTML-Tables with row-span / c

DaveP
  I'd go further, and request that processors warn users, even
  if they don't fall over or blow up or report an error.

That would be better than just warning in the spec, but without
completely ripping up xpath2 and starting over it's not really possible.

Most of the Backward Compatibility behaviour in xpath2 and xslt2 is
local: you set the BC flag and something acts a bit more like it did
in xslt1 within the scope of that setting.

But RTF usage is different it has a pervasive  nature,
You have
<xsl:variable name="x" vesion="1.0">
 <x/>
</xsl:variable>

The spec could define that this did not produce a document node but
instead (as in 1.0) produced something that was almost the same as a
document node, but could not be used with / or [].

The trouble is that that restriction would then have to be enforced for
the rest  of the scope of this variable, even in the bits of the
stylesheet not using BC mode, so that select="$x/x" made an error.

Thus this rtf type would have to be a full part of the Xpath 2 type
system, with its entries in the data model, formal semantics etc
this wouldn't be impossible but even I might admit this is asking for a
lot of specification work and implementation effort just to raise an
error.

David


-- 
http://www.dcarlisle.demon.co.uk/matthew

________________________________________________________________________
This e-mail has been scanned for all viruses by Star Internet. The
service is powered by MessageLabs. For more information on a proactive
anti-virus service working around the clock, around the globe, visit:
http://www.star.net.uk
________________________________________________________________________

 XSL-List info and archive:  http://www.mulberrytech.com/xsl/xsl-list