2008/7/1 Dan Price <[EMAIL PROTECTED]>:
> On Tue 01 Jul 2008 at 04:34PM, Shawn Walker wrote:
>> >> I plan on reviewing Dan's changes after work later tonight.
>> >
>> > utcnow() does not appear to actually add a UTC timezone to the datetime.
>> > I found the python docs for this stuff to be somewhat incomprehensible,
>> > so I decided that I would just assume all datetimes were in UTC, and
>> > take care to generate them in UTC.
>>
>> I'm not certain about that. I believe I found local timestamps when I
>> was working with the catalog for the RSS/Atom feeds.
>>
>> I'll look closer later tonight.
>
> The point I was trying to make (but obliquely) is that there seem to be
> two kinds of datetime: those which carry TZ info in them, and those that
> do not.  Calling datetime.datetime.now() or datetime.datetime.utcnow()
> both yield datetimes which do *not* carry a tzinfo.
>
> If we ensure that all of our datetimes are filled with UTC timestamps,
> then they don't need to have time zone info in them-- in other words,
> as long as you're consistent, and don't need to convert back and forth,
> it all pretty much works.  As Johansen pointed out, there are other
> spots in the code which need to be updated to be based on UTC time,
> and the catalog is one of them...

I agree, and yes it looks like catalog needs some fixes in this area.

Thanks,
-- 
Shawn Walker
_______________________________________________
pkg-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/pkg-discuss

Reply via email to