Hi David,

OK, that's really bad.  Can you provide more details about what we need to
reproduce this and we'll see what we can do?
I used

  http://www.sarvega.com/xslt-benchmark.php

to download the benchmark.

There is a pdf document describing the results and the testbed with all sample data and stylesheets. I have not compiled the driver for xalan, but used the testXSLT command line for testing.

"""
time `testXSLT17 \
      -IN testfiles/docbook-xsl/test/book2.xml \
      -XSL testfiles/docbook-xsl/html/docbook.xsl \
      > /dev/null  2> /dev/null`
"""
=>
real    0m3.728s
user    0m3.670s
sys     0m0.040s

"""
time `saxon testfiles/docbook-xsl/test/book2.xml \
     testfiles/docbook-xsl/html/docbook.xsl \
     > /dev/null  2> /dev/null`
"""
=>
real    0m2.553s
user    0m2.360s
sys     0m0.100s

It seems not so bad for xalan, but the saxon-script has to load the jvm and the real benchmark will be done with precompiled stylesheets.

HolgeR
--
holger floerke                      d  o  c  t  r  o  n  i  c
email [EMAIL PROTECTED]          information publishing + retrieval
phone +49 2222 9292 90              http://www.doctronic.de




Reply via email to