|
|
Boost Testing : |
From: Misha Bergal (mbergal_at_[hidden])
Date: 2005-06-01 00:52:55
Doug Gregor <dgregor_at_[hidden]> writes:
[...]
> I've been digging through this a bit, and realized that the "issues"
> page generator does essentially what we want... we'd just de-HTMLize
> it and sent it off to the list. Later, a post-processing step could
> determine just who would get the e-mail CC'd to them.
> However, I've been running the script to build the reports and the
> generation of issues.html has been going for a very, very long time
> (90 minutes)... is this why it hasn't been generated for the last few
> months?
Mostly because base on 1.32.0 experience we had some reservations
about its usefulness and thus it was not cost-effective to update it
with the rest of the reports. Now it probably is very not up-to-date.
> Perhaps we could optimize the XSLT enough to make it feasible again?
I think it can be easily done. Here are some ideas:
* the test data can be generated using
xsl_reports/test/generate_test_results.py
* I would make sure that issues.xsl generates correct results first.
* Then I would use a real regression logs and commenting out parts of
XSLT code to find out what the bottlenecks are.
Don't hesitate to ask more questions - we very much appreciate you
looking into this.
-- Misha Bergal MetaCommunications Engineering