David Carlisle <davidc@xxxxxxxxx> writes:

>>    They have 2 different specs and one can exist independently of the
>> other
>
> xpath can exist without xslt but not the other way round. The situation
> is (exactly) the same in XQuery, but XQuery is usually regarted as an
> extension of XPath: that is XQuery is a single language, with more
> constructs than XPath) whereas XSLT is usually described is a
> two-language construct consisting of xslt constructs and Xpath
> constructs. It's pretty much a marketing angle which way you describe it
> really. Although on the surface the XQuery spec doesn't defer to XPath
> for the specification of the Xpath-part of XQuery but rather just includes
> copies of the definitions, whereas the XSLT spec does refer to the xpath
> spec, this is just an artifact of the way the stylesheets making the
> public html versions of the spec are built. The XPath and XQuery
> documents are built out of a common xml document base.

I wouldn't mind if the only if was in xpath. In other words if there
were *more* ties that bind between the two worlds:

  <xsl:variable name="x" 
    select="if $var then select(@id='1') else select (@id='2')"/>
  <something id="1">
     hello!
  </something>
  <otherthing>
     goodbye
  </otherthing>

If we are to have the facilities in xpath I can't see why they can't
point into the xslt structure.

Ah well. I should have joined the working group (but I couldn't afford
it!)


-- 
Nic Ferrier
http://www.tapsellferrier.co.uk   for all your tapsell ferrier needs