Re: [Boost-testing] Meaning of the date in detail regression report?

From: Aleksey Gurtovoy (agurtovoy_at_[hidden])
Date: 2006-12-15 02:39:46

Roland Schwarz writes:
> Aleksey Gurtovoy wrote:
> >> 1) The date the snapshot was taken
> > 
> > That's the timestamp displayed under the runner's name.
> > 
> >> 2) The date the test had been run
> > 
> > This one is available on the test's output page, but unless specified
> > otherwise, the page is produced only if the test fails.
> > 
> 
> The time stamps still are suspect:
> 
> E.g.:
> http://engineering.meta-comm.com/boost-regression/CVS-RC_1_34_0/developer/thread_release.html
> 
> Report Time: Thu, 14 Dec 2006 13:54:52 +0000
> 
> CVS Update time: RudbekAssociates-V2 Thu, 14 Dec 2006 12:15:00 +0000
> 
> Compiler output [2006-12-02 13:50:41 UTC].
> 
> What is strange here is, that the CVS Update is _later_ than the 
> compiler output!

This a normal situation if the regressions are run in the incremental
mode (or using a custom script that does something equivalent to the
incremental mode, which, I believe, is the case with Victor), when the
tests are re-run only when their dependencies have changed.

-- 
Aleksey Gurtovoy
MetaCommunications Engineering