RE: What will be the future improvements of XSLT?

-----Original Message From: Tangi Vass [mailto:tangivass@xxxxxxxxxxxxxx]

>However, I'm very disappointed by the 'weaknesses' of the current version
of XSLT, mainly >the following two:
>  the non-mutability of variables (xsl:assign is not part of the standard),

>  loops and templates can't work on, nor return, result tree fragments
(forbidding multiple passes, at least without dirty hacks).


Yes, a lot of people find programming in XSL difficult at first. I certainly
did.

The non-mutability of variables is a conceptual hurdle especially for
"jobbing programmers" who have perhaps only been exposed to procedural
languages in the past. It is less of a barrier, in my view, for
non-programmers and for those with a computer science training. But I
personally believe the benefits of this design principle do not justify the
restrictions it imposes, which is why I have implemented assignment (and
loops) in SAXON.

I think that a function to convert a result tree fragment into a node-set
that can be further processed is a perfectly reasonable extension to the
standard and I've started experimenting to see if it can be done in SAXON.

Our main concern in introducing XSL into various real live customer projects
is whether we can train the graphical designers and editorial staff to use
it. The real problem here is that XSL is far too rich in concepts for people
accustomed to "HTML with CSS stylesheets and a little JavaScript".
Unfortunately I don't think the review process in W3C includes usability
trials. Perhaps we will have to write some front-end tools to make it
viable.

Mike Kay


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