This is the mail archive of the
xsl-list@mulberrytech.com
mailing list .
Re: ANNOUNCE: Petition to withdraw xsl:script from XSLT 1.1
- To: Jeni Tennison <mail at jenitennison dot com>
- Subject: Re: [xsl] ANNOUNCE: Petition to withdraw xsl:script from XSLT 1.1
- From: "Clark C. Evans" <cce at clarkevans dot com>
- Date: Fri, 2 Mar 2001 05:50:13 -0500 (EST)
- cc: xsl-list at lists dot mulberrytech dot com, RobinBerjon <robin at knowscape dot com>
- Reply-To: xsl-list at lists dot mulberrytech dot com
On Fri, 2 Mar 2001, Jeni Tennison wrote:
> > I had in mind a RDDL [1] like mechanism where one could ask for a
> > "your-favorite-language" implementation of the "some-uri-reference"
> > extension function. Certainly this means going out to the 'net for
> > implementations. I include by reference all of the caching
> > discussions relted to RDDL.
>
> I really like the RDDL idea, especially as you've expanded it in later
> messages. I love the language-neutral function description. I also
> think the indirection will discourage people from just 'opting out' to
> Javascript etc. when they can't see how to achieve something in XSLT.
Thanks Jeni, the language-neutral module specification and binding
follows naturally from the following comment made on xml-dev:
On Thu, 1 Mar 2001, Robin Berjon wrote:
> My quarrel with xsl:script is that it *immediately* ties an extension to
> an implementation language. This implies either that all implementations
> will have to support the same (set of) languages and thus resort to the
> usual LCD of programming languages (ecmascript) or that stylesheet
> authors will have to include xsl:scripts of equivalent functionality in
> several different languages with every stylesheet that they want to be
> portable. That's just insane.
It is Robin who identified that the primary problem with xsl:script
is that the association between the identifier and the implementation
code is immediate. Is just adding a layer of indirection all that
is needed? It's nice that the language-independent module identifier
can also be used as an RDDL locator...
The other neat thing about adding an extra layer of indirection is
that this "binding" stuff can all be done is a seperate specification
independent of XSLT. This would allow the same mechanism to be
used in other specifications...
Clark
XSL-List info and archive: http://www.mulberrytech.com/xsl/xsl-list