Xquery for 'hard data probs' was RE: [xsl] XSLT vs Perl

Subject: Xquery for 'hard data probs' was RE: [xsl] XSLT vs Perl
From: "Jim Fuller" <jim.fuller@xxxxxxxxxxxxxx>
Date: Fri, 20 Feb 2004 19:52:43 -0000

>I don't quite understand how you come to these conclusions.

Apologies, missed this email.

>XQuery 1.0 is a simpler language than XSLT 2.0, without doubt. Its
>functionality is roughly equal to XSLT 1.0 minus xsl:apply-templates,
>plus xsl:function. This isn't meant as a criticism. But I find it

Quite a succint distillation of the differences, it's a bit
disconcerting that the differences between Xquery and XSLT can be boiled
down into 1 paragraph though !

>difficult to see why you think it is capable of tackling harder
>than you would tackle with XSLT.

I see XQuery as the 'SQL of the xml world', it is applied to querying
xml data. XSLT is designed for transformation of xml. In reality, the
differences are few and far between....my own usage of XSLT has forced
me to create more optimised data structures...effectively reducing the
complexity of my data problem...which mark some of the 'hard data
problems' I quess I was thinking of....the old hack is 'complicated
joins' but this is just iteration of complexity...not magnitude jumps in
the complexity. I guess I may have just been spouting nonsense re 'hard
data problems', but alternately I don't consider it bad advice to
suggest using Xquery for querying data.

to me the 'hard data problems' was probably more in reference to typical
SQL situations e.g. those which require complicated multiple joins,
triggers, preserving data integrity, managing transactions and roll back
across multiple data domains.... all of which can have an impact on
performance and stability if not designed and performed correctly (read
lots of specialist knowledge to do right). This type of thing I quess is
always outside the scope of both XSLT and Xquery to solve....something I
find frustrating; I quess I would like to see more 'typical scenarios'
which represent 'real' problems instead of pristine use cases....the
real problems in computing typically cluster around the boundaries and
interfaces of a technology.

But you have proved your point to me, there is no difference between
using XQuery or XSLT in solving any data problem (hard or otherwise), so
why are there 2 languages ? I guess if there were any XQuery use cases
that XSLT could not perform I would feel more comfortable with 2
languages existing, but that is a moot point. In any event, for
procedural types Xquery will no doubt prove popular. 

I would be interested in your opinion if Xquery is more superior in
querying data over XSLT, perhaps u have some clear use cases gathered
whilst developing your Xquery implementation.

Cheers, Jim Fuller

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

Current Thread