Re: Bug #630: date/time storage problem: timestamp parsed

Sean Chittenden wrote:

>>>The FreeBSD folk are absolutely adamant about having mktime() no
>>>compensate for deadzones between DST shifts and they insist that
>>>the application handle this. Someone's off looking at how other
>>>OS'es handle this, but this could be an arduous battle on that
>>>front. <:~)
>>>
>>Personally I'd like to see FreeBSD do away with this strange
>>behaviour. It cause my grief because certaint hings *MUST* be done
>>at 0200 every day in our system, I was forced to do them manually
>>recently, shifing several hours of work into daytime which had to be
>>paused and bulked into the next days work. I realise that this is
>>getting off track but it just points out that the FreeBSD behaviour
>>is IMHO WRONG. It causes applications to fail in an unexpected and
>>odd way.
>>
>>I'm not objecting to pg patching for it (no choice at the moment)
>>but I hope the pg team 'officially' puts a little pressure on the
>>BSD folk to make this behave as expected.
>>
>
>Feel free to read over their arguments (archive may not be 100% up to
>date):
>
>http://docs.freebsd.org/mail/archive/2002/freebsd-standards/20020414.freebsd-standards.html
>
>>I don't have any compliance docs at the moment, but this strikes me
>>as somewhat out of spec personally.
>>
>
>::shrug:: I've gotten enough push back to have an indifferent opinion:
>I just want to see PG work w/ some of the bogus data I get every now
>and then. :~) -sc
>
Yes but not everyone changes over at 2AM on the specific day. The rest
of the world for the most part doesn't in fact. I don't know what
mktime() behaviour is in different locales (IE different TZs) but if its
the same (IE deadzone @ the same time when the TZ is something in say
the EU who follow different rules) then its broken. I've got a FreeBSD
4.3 box here I do most of my serving on I'll see if I can get a little
time to do some testing with different TZs. I don't think that the way
BSD handles it is correct

Also browsing the discussion archives it seems that mktime() atleast on
BSD is inconsistent with how it handles bogus dates anyway. Looks like
the BSD guys are going to be doing a little navel-looking over this.