RE: XSLT 1.1 status (was [xsl] document.write)

["Michael Kay" <michael dot h dot kay at ntlworld dot com>]

> I understand, and there's been a lot of bickering about
> xsl:script.  But
> when I read the XSLT 2.0 requirements (at
> http://www.w3.org/TR/xslt20req
> )and see
>
> "Any stylesheet whose behavior is fully defined in XSLT 1.1 and which
> generates no errors will produce the same result tree under XSLT 2.0"
>
> --then I assume xsl:script will be in 2.0--as will everything
> else in 1.1 WD.

We (the W3C XSL WG) should have updated the requirements document to change
"1.1" to "1.0". The XSL Working Group is not committed to retaining XSLT 1.1
syntax in XSLT 2.0 in a compatible form. In fact this applies to any working
draft, it would have been true even if XSLT 1.1 had been progressed.

However, we are treating all the functionality of XSLT 1.1 as being on the
requirements list for XSLT 2.0, with one notable exception: the
standardization of Java and JavaScript language bindings. The WG has made a
decision that is best kept out of the core XSLT specification: some members
were hoping that the Java bindings might be picked up in a future JAXP
specification instead.

> I guess I'm not clear on what "defined in XSLT 1.1 and
> which generates
> no errors" means, since whether it generates errors depends on your
> implementation--and as you've said, xsl:script is not widely
> supported.

The qualification (if you change 1.1 to 1.0) means that the compatibility
requirement only applies to stylesheets that contained no errors as far as
the XSLT 1.0 specification is concerned. If a stylesheet produced an error
in 1.0, it's not guaranteed to produce the same error in 2.0. (For example,
select="$rtf//item" might start working...). This has nothing to do with
implementations - the spec determines whether a stylesheet is in error, not
an implementation.

Mike Kay


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