Re: os.time() and timezones

[Daurnimator <quae@...>]

On 14 April 2012 06:26, Sean Conner <sean@conman.org> wrote:
>  Even operating systems have trouble with timezones:
>
> http://www.chronos-st.org/Discovering%20the%20Local%20Time%20Zone--Why%20It%27s%20a%20Hard%20Problem.html
>
>> I use this kind of logic to find the timezone offset in pl.Date:
>>
>>         local t = os.time()
>>         local ut = os.date('!*t',t)
>>         local lt = os.date('*t',t)
>>         thour = lt.hour - ut.hour
>>         tmin = lt.min - ut.min
>>
>> If this is misguided I would certainly like to know!
>
>  It is.  Try this:
>
> gmt = {  0 ,  1 ,  2 ,  3 ,  4 ,  5 ,  6 ,  7 ,  8 ,  9 , 10 , 11 ,
>        12 , 13 , 14 , 15 , 16 , 17 , 18 , 19 , 20 , 21 , 22 , 23 }
> loc = { 19 , 20 , 21 , 22 , 23 ,  0 ,  1 ,  2 ,  3 ,  4 ,  5 ,  6 ,
>         7 ,  8 ,  9 , 10 , 11 , 12 , 13 , 14 , 15 , 16 , 17 , 18 }
>
> for i = 1 , 24 do
>  print(i,loc[i] - gmt[i])
> end
>
> And even when it does work, it still doesn't take into account DST.  I found
> this works:
>
> now   = os.time()
> lmt   = os.date("*t",now)
> gmt   = os.date("!*t",now)
> timel = os.time(lmt)
> timeg = os.time(gmt)
> zone  = os.difftime(timel,timeg)
>
> if lmt.isdst then
>  if zone < 0 then
>    zone = zone + 3600
>  else
>    zone = zone - 3600
>  end
> end
>
>  -spc (Time is hard!  I wonder how The Doctor does it?)
>
>

And now you're assuming daylight savings is for an hour.
Also that timezones are only on the hour.

eg; see Lord Howe island in Australia: LHDT/LHST