Bug #118694 for Time-Zone-Olson: Using date(1) in tests

Thu Nov 10 07:08:38 2016 andreas.guldstrand [...] gmail.com - Ticket created

Subject:	Using date(1) in tests
Date:	Thu, 10 Nov 2016 13:08:29 +0100
To:	bug-Time-Zone-Olson [...] rt.cpan.org
From:	Andreas Guldstrand <andreas.guldstrand [...] gmail.com>

Using date(1) in the tests will not work on for example Windows, even if they've got themselves a /usr/share/zoneinfo database. I'm unsure about a way to to run the same tests without it though. Perhaps skip them if the date program doesn't give reasonable output when you do "date --version"?

Fri Nov 11 03:46:38 2016 DDICK [...] cpan.org - Correspondence added

On Thu Nov 10 23:08:38 2016, andreas.guldstrand@gmail.com wrote: Show quoted text

> Using date(1) in the tests will not work on for example Windows, even > if they've got themselves a /usr/share/zoneinfo database. > > I'm unsure about a way to to run the same tests without it though. > Perhaps skip them if the date program doesn't give reasonable output > when you do "date --version"?

Yeah, I usually make an effort to code for the Win32 platform. This module has been an exception so far. There is probably a WMI interface i could interogate to test timezone information on Win32.

Fri Nov 11 03:46:38 2016 The RT System itself - Status changed from 'new' to 'open'

Tue May 28 16:41:47 2019 DDICK [...] cpan.org - Correspondence added

On Fri Nov 11 19:46:38 2016, DDICK wrote: Show quoted text

> There is probably a WMI interface i could interogate to test timezone > information on Win32.

I wound up using the Win32 API call SystemTimeToTzSpecificLocalTime

Tue May 28 16:41:48 2019 DDICK [...] cpan.org - Status changed from 'open' to 'resolved'

Tue May 28 16:41:48 2019 DDICK [...] cpan.org - Fixed in 0.18 added