RE: Comparing grouping techniques in terms of performa

OK Kevin, thanx for the pointer - looks good. After equalizing my
stylesheets I found that the template approach was slightly faster, but I
also saw that the times varied a lot from one transformation to another. The
good thing is that the big difference in execution time was just caused by
myself copying in a double <xsl:key> declaration - I still ask myself how
did it come there :-)

<prs/>
http://www.pietsieg.com
http://www.pietsieg.com/dotnetnuke
Contributor on www.ASPToday.com
Co-author on "Professional ASP.NET XML with C#", July 2002 by Wrox Press

-----Original Message-----
From: Kevin Jones [mailto:kjones@xxxxxxxxxxx] 
Sent: Wednesday, April 07, 2004 2:40 AM
To: xsl-list@xxxxxxxxxxxxxxxxxxxxxx
Subject: Re:  Comparing grouping techniques in terms of performance

On Tuesday 06 April 2004 22:39, Pieter Reint Siegers Kort wrote:
>
> I still would like to know if there are any sites that publish this 
> type of comparison - if not, I would be interested in setting up this 
> myself as part of my site.
>

Hi,

There are unfortunately very few published performance guidelines. The best
overview is probably in the FAQ at
http://www.dpawson.co.uk/xsl/sect4/N9883.html. 

I think it is fairly well understood that all others things being equal
using for-each will be quicker than template matching. The caveat is that
the improvement is often so small as not to be important and template
solutions are normally much cleaner. In this particular case (with
simplified stylesheets) I measured a 4% difference for a file with 2000
entries in favour of using for-each. As the document size grows so does the
difference. 
This was on the Sarvega processor, other processors may be different.

Regards,
Kev.