Re: [xsl] XSLT 2.0 or XSLT 1.0 -- which is more elegant?

[Eric van der Vlist <vdv@xxxxxxxxxxxx>]

Hi Dimitre,

On mar, 2005-04-19 at 20:23 +1000, Dimitre Novatchev wrote:

> With XSLT 2.0 I can have expressions like:
> 
>     f:pow(sum(f:map(f:flip(f:pow(),10), 1 to 10)), 0.1)
> 
> or
> 
>     f:transform-and-sum(f:flip(f:substring-before(), '*'), 
>                                        data(/*/*/@colwidth))
> 
> 
> or
> 
>     f:map(f:round-half-to-even(f:sqrt(2, 0.000001)), 0 to 13)
> 
> 
> These are really elegant compared to the xslt 1.0 code necessary to
> produce the same result.
> 
> Another reason I find XSLT 2.0 more elegant than XSLT 1.0 is that
> there isn't anymore any need to use an xx:node-set() extension
> function.
> 
> What is really not elegant at all in XSLT 2.0 is the impossibility to
> define user data types inline in a stylesheet  -- forcing the
> programmer to artificially separate in different files type definition
> from type usage makes XSLT 2.0 rather unique... :(
> 
> Of course, one would also like to see nested sequences, type classes
> and type equations...

I won't argue that XSLT 2.0 doesn't bring useful features and I must
admit I am using it punctually, but I consider that the flaws of the
PSVI based architecture promoted by XSLT 2.0 generally outweigh the
benefit of these features and that's what prevent me from using it on a
large scale.

Of course, that's a decision which is very subjective and even personal
since it's a trade-off between architectural principles and concrete
features :-) ...

Eric
-- 
Carnet web :
           http://eric.van-der-vlist.com/blog?t=category&a=Fran%C3%A7ais
------------------------------------------------------------------------
Eric van der Vlist       http://xmlfr.org            http://dyomedea.com
(ISO) RELAX NG   ISBN:0-596-00421-4 http://oreilly.com/catalog/relax
(W3C) XML Schema ISBN:0-596-00252-1 http://oreilly.com/catalog/xmlschema
------------------------------------------------------------------------