Date: Wed, 29 May 1996 12:14:13 +0200 (IST)
From: Eli Zaretskii <eliz AT is DOT elta DOT co DOT il>
To: Clive Page <cgp AT le DOT ac DOT uk>
Cc: djgpp AT delorie DOT com
Subject: Re: What is all this "timezone" stuff for?
In-Reply-To: <4oehjq$o95@hawk.le.ac.uk>
Message-Id: <Pine.SUN.3.91.960529120624.12693L-100000@is>
Mime-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII


On 28 May 1996, Clive Page wrote:

> as was).  I have to say that this simply isn't possible.  The duration of
> Summer Time (as it's called in the Northern hemisphere) varies from place
> to place quite randomly and unpredictably, and changes are at the whim of
> governments, which don't have any sense of needing to conform with
> algorithms or standards.  Recently lots of diaries and calendars

That's why you get the source of these files and a compiler that 
generates the binary zoneinfo files.  Just change the source relevant to 
your country and recompile it with zic, or set the TZ variable to a full 
definition of your zone, and you are done.

> I think the whole idea of predicting the local-time - UTC difference is
> fatally flawed.  It seems silly for the DJGPP package to even attempt it.
> My respect for the sagacity of those who invented the Posix series of
> standards, already not all that high, has fallen further, now I know that
> they are responsible for clogging my disc with useless junk.

If this junk is ``useless'' for you, just delete it and use DJGPP as if 
you were in GMT0 zone (which seems to be quite close to your time zone 
anyway).  Any DJGPP user who isn't connected to a network which spans 
time zones, and doesn't bring zip files from other time zones (or doesn't 
care about time stamp incompatibilities), can safely pretend that they 
are in GMT0, which actually means the times reported by `stat' are local 
times.

But there *are* people who do care, and since all the other machines
connected to the net use this convention, there is no other way for them 
but to comply.