[Mailman-Developers] Timestamps in archive are +1 hr

[John W. Baxter]

Version of Mailman?  System on which you're running?

Assuming Linux, what is the third (last) line of
cat /etc/adjtime

I expect either LOCAL or UTC...our Mailman machine has UTC, and--which I had
never noticed--is an hour ahead on the From lines in the archive.  Hmmm.

Mailman 2.1.2; RedHat 9

  --John


On 8/17/2004 2:12, "Chris Boulter" <chris at jellybaby.net> wrote:

> Apologies for noise, but I posted this to mailman-users last month and got
> no replies. It doesn't seem like something which would be all that hard
> to fix for someone who knows Python/Mailman better than I do. Any ideas?
> 
> ----- Forwarded message from Chris Boulter <chris at jellybaby.net> -----
> 
> Date: Wed, 28 Jul 2004 15:30:35 +0100
> From: Chris Boulter <chris at jellybaby.net>
> Subject: [Mailman-Users] Timestamps in archive are +1 hr
> To: mailman-users at python.org
> 
> Hi,
> 
> The timestamps on messages in our Mailman archives are all 1 hour ahead of
> reality. Any ideas?
> 
> I'm in the UK, where it's currently summertime, which is UTC +0100. Typing
> 'date' on our Mailman box gives
>       Wednesday July 28 15:24:33 BST 2004
> which looks correct to me.
> 
> I've found that I can fix the time by changing i18n.py line 60 (in the
> ctime function) from
>       year, mon, day, hh, mm, ss, wday, yday, dst = time.localtime(date)
>                                                          ^^^^^^^^^
> to
>       year, mon, day, hh, mm, ss, wday, yday, dst = time.gmtime(date)
>                                                          ^^^^^^
> and rebuilding the archives, but I don't want to leave this hack in place.
> 
> If it's useful to know, a test mail I sent at 1253 local time today had
> its seconds-since-epoch set to 01091019325 by the time it got to the
> archive.
> 
> Many thanks.