On 22/04/12 05:49, Tom Lane wrote:
> eshkinkot@gmail.com writes:
>> set timezone to 'W-SU'; select '2011-03-27 23:00:00'::timestamptz;
>> SET
>> timestamptz
>> ------------------------
>> 2011-03-28 02:59:54+04
> Bizarre. On my Fedora 16 box, I see a different misbehavior:
>
> regression=# set timezone to 'Europe/Moscow'; select '2011-03-27 23:00:00'::timestamptz;
> SET
> timestamptz
> ------------------------
> 2011-03-28 00:29:40+04
> (1 row)
>
> (W-SU behaves the same, incidentally.) Two other machines are fine
> with this case, though. It may or may not be relevant that the F16
> build is using --with-system-tzdata, as I imagine your Debian package
> is also, while the machines that are happy are not.
>
> So: some platform-specific misbehavior here. I have no time to poke at
> it more now, though. Who else can reproduce this, on what platforms?
>
> regards, tom lane
>
I get strange results on Fedora 16 (I think I upgraded straight from 14)!
$ psql
psql (9.1.3)
Type "help" for help.
gavin=> set timezone to 'Europe/Moscow'; select '2011-03-27
23:00:00'::timestamptz;
SET
timestamptz
------------------------
2010-05-02 05:46:24+04
(1 row)
gavin=> \q
$ uname -a
Linux saturn 3.3.2-1.fc16.x86_64 #1 SMP Sat Apr 14 00:31:23 UTC 2012
x86_64 x86_64 x86_64 GNU/Linux
$ date
Sun Apr 22 09:27:10 NZST 2012
$