[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [MiNT] TZ change?



Hi,

On Sun, Oct 31, 1999 at 06:36:04AM +0200, Martin-Eric Racine wrote:
> On Tue, 26 Oct 1999, Guido Flohr wrote:
> 
> > > I noticed this morning that ircII shows a time that is one hour
> > > ahead of the system time.  Since it was compiled in USA (Kellis,
> > > TZ:ESTEDT) I was wondering if USA just changed to Winter time?
> 
> And it is now back to normal, since tonight.

I'm a prophet, you see. ;-)

> It switched to EET, presumably at midnight, but not all running
> processes noticed. For instance, inetd (port 37:time, internal)
> returned "GMT+3" instead of "GMT+2" when using rdate from Stacy.

At 2 o'clock, not at midnight.

> > I think they change on the last Sunday in October, not sure.
> 
> Do we also, in Europe?
> 
> As I remember, Europe uses this "last Sunday of..." rule, both
> when changing to Winter time and again back to Summer time, but
> Canada/US don't use the same simple method.

That's what the tz database is for.  Everything is stored there.

> Anyhow, my point is, binaries should change timezone according
> to the machine on which they're running, not to whoever ported
> some package and _their_ timezone.

So what? Do you think this is not the case?

> > 	zic -l Europe/Helsinki -p Europe/Helsinki
> 
> That was done when I installed the newest tzinit from mintlib
> 0.53 and most packages seem to recognize this zone if they were
> compiled with MINTLIB 0.49 or newer.  
> 
> MiNTOS is among those that need to be recompiled, otherwise login
> and syslog produce incorrect timestamps (RE: my previous email,
> when the first version of tzinit was released about a year ago -
> mine was rebuilt using MINTLIB 0.49 back then).

You can avoid that by passing an appropriate TZ envariable to old programs
in your startup scripts.  However you have to change that manually
whenever you change the timezone.

Ciao

Guido
-- 
http://stud.uni-sb.de/~gufl0000/
mailto:gufl0000@stud.uni-sb.de