[rb-general] TZ=UTC wrong?

Bastian Blank waldi at debian.org
Wed Oct 30 13:20:11 UTC 2019


Hi Holger

On Wed, Oct 30, 2019 at 12:42:54PM +0000, Holger Levsen wrote:
> On Tue, Oct 29, 2019 at 09:00:15PM +0200, Niko Tyni wrote:
> > On Mon, Oct 28, 2019 at 12:28:32PM +0100, Guillem Jover wrote:
> > > Just noticed this change from the changelog. :) UTC is not really a
> > > proper timezone specification, the format requires an offset, so here
> > > it would be UTC0 (see «man timezone»).

Those are two different concepts.  UTC refers to a timezone file,
/usr/share/zoneinfo/UTC.  There are several files called UTC:

| % find /usr/share/zoneinfo -name "UTC*" 
| /usr/share/zoneinfo/Etc/UTC
| /usr/share/zoneinfo/posix/Etc/UTC
| /usr/share/zoneinfo/posix/UTC
| /usr/share/zoneinfo/right/Etc/UTC
| /usr/share/zoneinfo/right/UTC
| /usr/share/zoneinfo/UTC

UTC0 or UTC+0 defines a unknown timezone called UTC and is 0 hours ahead
of the default UTC.

Just compare those:

| % TZ=UTC+0 date
| Mi 30. Okt 12:58:54 UTC 2019
| % TZ=UTC-1 date
| Mi 30. Okt 13:58:56 UTC 2019
| % TZ=UTC+1 date
| Mi 30. Okt 11:58:58 UTC 2019

Or even:

| % TZ=WAT+0 date  
| Mi 30. Okt 13:00:34 WAT 2019

> What's the opinion of debian-devel@ and others?

If you don't want to use the info in the zoneinfo files, specify UTC+0,
or DEB+0, or WAT+0.

Aren't those localtime maximum and minimum values bogus?  The rely on
the timezone during build, not during execution.  General speaking it is
only safe to use values between -2^63+1day and 2^63-1day or so for 64bit
time_t anyway.

Regards,
Bastian

-- 
Violence in reality is quite different from theory.
		-- Spock, "The Cloud Minders", stardate 5818.4


More information about the rb-general mailing list