[Buildbot-commits] [Buildbot] #2362: Timezone problem with HgPoller

[Buildbot]

#2362: Timezone problem with HgPoller
---------------------+--------------------
Reporter:  gracinet  |       Owner:
    Type:  defect    |      Status:  new
Priority:  major     |   Milestone:  0.8.7
 Version:  master    |  Resolution:
Keywords:            |
---------------------+--------------------

Comment (by gracinet):

 Thanks for the quick reply.
 I'm sorry to bother you with that, but I'm really puzzled there, because
 what you describe is what I implemented.
 I start believing the web UI (the waterfall, notably) is actually wrong.

 Here's a changeset I made yesterday, seen from the web UI (from changeset
 details page)::

 {{{
 Changed at      Tue 28 Aug 2012 12:57:44
 Revision        ee779e43f8e0e5af90a318732392c28fe60c1153
 }}}

 That time is actually correct, but is expressed in UTC. The original time
 was CEST ::

 {{{
 $ hg log -r ee779e43f8e0e5af90a318732392c28fe60c1153 | grep date
 date:        Tue Aug 28 14:57:44 2012 +0200
 }}}

 For the record, the log used by HgPoller is
 {{{
 hg log -r ee779e43f8e0e5af90a318732392c28fe60c1153
 --template='{date|hgdate}\n'
 1346158664 -7200
 }}}

 !HgPoller sums the two values, to get the UTC timestamp. This seems also
 correct.

 On the changeset details page, it would be better to localize to the
 buildmaster timezone (CEST again), since users don't expect an UTC time,
 but it's a minor detail.

 The real usability problem is with the waterfall, whose timeline column
 claims to be in CEST (that's what I have in the column head). But the
 changeset appears at 12:57 in the timeline instead of the expected 14:57.
 The subsequent build appears at 14:57 in the waterfall (correct), and
 looks unrelated to the changeset that triggered it.

 I'll take a look at the web UI code and keep you posted.

-- 
Ticket URL: <http://trac.buildbot.net/ticket/2362#comment:2>
Buildbot <http://buildbot.net/>
Buildbot: build/test automation