Re: [PATCH 1/3] Move Git::SVN::get_tz to Git::get_tz_offset
- From
Junio C Hamano <gitster@pobox.com>
- Date
- Jan 16, 2013, 20:36 UTC
- Message-ID
- <7vehhkx3ul.fsf@alter.siamese.dyndns.org>
- In-Reply-To
- <CAP30j164UD9gNRbZ=uCQjgpDODWnGtYmHcWES2P=YPryL=FbZA@mail.gmail.com>
Ben Walton <bdwalton@gmail.com> writes:
Show 18 quoted lines
> On Wed, Jan 16, 2013 at 3:37 PM, Junio C Hamano <gitster@pobox.com> wrote:
>> Ben Walton <bdwalton@gmail.com> writes:
>>
>>> +sub get_tz_offset {
>>> + # some systmes don't handle or mishandle %z, so be creative.
>>
>> Hmph. I wonder if we can use %z if it is handled correctly and fall
>> back to this code only on platforms that are broken?
>
> That would be perfectly acceptable to me. The reason I set it up to
> always run through this function here is that when I originally added
> this function for git-svn, I'd made it conditional and Eric Wong
> preferred that the function be used exclusively[1]. I opted to take
> the same approach here to keep things congrous.
>
> If it were to be conditional, I think I'd add a variable to the build
> system and have the code leverage that at runtime instead of the
> try/except approach I attempted in 2009.If the code was originally unconditional for a reason (and I think being bug-to-bug compatible across platforms is actually a good thing in a tool like importers), I would not object to it. Thanks for the back-story.