Re: [xsl] Re: namespace values

[Jeni Tennison <mail@xxxxxxxxxxxxxxxx>]

Hi Dimitre,

> The EXSLT initiative is a good initial set of user requirements for
> ***the real thing to come in XSLT 2.0***, but not more than that.

User requirements generally don't give solutions to the requirements,
and EXSLT does. I certainly hope that the discussions we have now
about the functionality of the various parts of EXSLT will have some
input into the design of XSLT 2.0 and XPath 2.0, but XSLT 2.0 and
XPath 2.0 already have sets of requirements.  The intention of EXSLT
is certainly not as a requirements document.

> Anybody has the right to invent a totally new language for
> processing XML documents. The danger is not to really start thinking
> this is standard XSLT and mixing the two together.
>
> The just published "rationale" proves these concerns:
>
> "XSLT processor implementers should implement the extensions as
> documented, or incorporate third-party implementations of the
> elements and functions."
>
> Nobody can force an implementor to implement/incorporate a set of
> extensions. EXSLT does not have the normative force of a W3C
> Recommendation.

What (official or unofficial) organisation you choose to trust to
define the functions you use is completely up to you. If you want to
stick with pure XSLT then no one's going to stop you, and I'm
certainly not planning to take a baseball bat to XSLT UK and beat
implementers into adopting EXSLT. Though I might bribe them with a few
drinks ;)

I did wonder about whether the 'should' was too strong, but then I
wondered what the point was of trying to draw together a standard (not
standard XSLT, I hope no one's making that misinterpretation, but a
standard definition of a set of extensions for use with XSLT) unless
we treat it as something that we can expect implementers to implement.

As Mike said, implementers respond to pressure from the people using
their software. If I had written 'can if they really feel like it' I
don't think any implementer would feel much pressure to do so. And if
they don't implement them, then there's absolutely no point to EXSLT -
you can't have something that's portable if it's only available in one
processor, as the multiple extensions in various processors
demonstrate. So that was why I chose that wording, but please give me
an alternative that makes you feel better about it and I'll use that
instead - as I said, that was a first attempt and I'd like to refine
it to something that everyone finds acceptable and rational.

I can't point to messages, but there have been several pleas for a
central library of standard extension elements and functions. EXSLT is
a response to that. If we misheard, and actually people don't care
about portability or having that centralised repository, then great,
we can stop wasting everyone's time. If it's just me trying to
organise it that you don't like, then fine, I'm very happy to hand it
over to someone else. If it's the content of the documents that you
find objectionable, then please say which bits and we can work on
changing them.

Cheers,

Jeni

---
Jeni Tennison
http://www.jenitennison.com/



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