Thread: Timezones -- what comes out does not go in?
The timezone on our boxes is set to Etc/UCT (possibly a distro default). (psql 8.1.4) => select to_char(current_timestamp, 'HH24:MI:SS TZ'); to_char -------------- 09:05:48 UCT (1 row) => select '09:05:48 UCT'::time; ERROR: invalid input syntax for type time: "09:05:48 UCT" UTC works, of course: => select '09:05:48 UTC'::time; time ---------- 09:05:48 (1 row) Is this behaviour expected? Desirable? Thanks Julian
"Julian Scarfe" <julian@avbrief.com> writes: > The timezone on our boxes is set to Etc/UCT (possibly a distro default). > (psql 8.1.4) > => select to_char(current_timestamp, 'HH24:MI:SS TZ'); > to_char > -------------- > 09:05:48 UCT > (1 row) > => select '09:05:48 UCT'::time; > ERROR: invalid input syntax for type time: "09:05:48 UCT" This works in CVS HEAD, thanks to some hard work by Joachim Wieland. http://archives.postgresql.org/pgsql-patches/2006-07/msg00077.php http://archives.postgresql.org/pgsql-committers/2006-07/msg00258.php regards, tom lane
Julian Scarfe wrote: > The timezone on our boxes is set to Etc/UCT (possibly a distro default). > > (psql 8.1.4) > > => select to_char(current_timestamp, 'HH24:MI:SS TZ'); > to_char > -------------- > 09:05:48 UCT > (1 row) > > => select '09:05:48 UCT'::time; > ERROR: invalid input syntax for type time: "09:05:48 UCT" > > Is this behaviour expected? Desirable? It is expected, because back in 8.1 the timezones recognized came from a hardcoded table. It's not desirable of course. But on the other hand, it's fixed in the development code: alvherre=# set timezone to 'Etc/UCT'; SET alvherre=# select to_char(current_timestamp, 'HH24:MI:SS TZ'); to_char -------------- 14:19:05 UCT (1 fila) alvherre=# select '14:19:05 UCT'::time; time ---------- 14:19:05 (1 fila) -- Alvaro Herrera http://www.CommandPrompt.com/ PostgreSQL Replication, Consulting, Custom Development, 24x7 support
From: "Tom Lane" <tgl@sss.pgh.pa.us> > This works in CVS HEAD, thanks to some hard work by Joachim Wieland. One of these days I'll find an issue *before* you folks have patched it. :-) Thanks Julian