[darcs-users] darcs benchmarking on buildbots

[Eric Kow]

Hi Karel,

I'm answering because Zooko mentioned that he would be very busy
this week.

On Wed, Sep 10, 2008 at 18:26:11 +0200, Karel Gardas wrote:
> I see darcs benchmarking causes a little side effect of not running
> darcs testsuite on majority of darcs' buildbots. May I ask (if it is
> possible) to move benchmarking after running all testsuite tests? (ie.
> after make test; make test_network).

I think I understand what is happening.  The test suite results are
buried deep in the waterfall:

 - benchmark results (compare with old darcs)
 - compile old darcs
 - fetch old darcs
 - test new darcs <--- here are the test results!
 - compile new darcs
 - fetch new darcs

Is this the case for you?  If so, perhaps we can re-order things as
you suggest:

 - benchmark results (compare with old darcs)
 - test new darcs <--- moved up so you can see it!
 - compile old darcs
 - fetch old darcs
 - compile new darcs
 - fetch new darcs

Thanks,

-- 
Eric Kow <http://www.nltg.brighton.ac.uk/home/Eric.Kow>
PGP Key ID: 08AC04F9
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
Url : http://lists.osuosl.org/pipermail/darcs-users/attachments/20080910/66e091fb/attachment-0001.pgp