proposed changes to TZ tables

Peter Ilieve peter at
Tue Dec 6 19:30:00 UTC 1994

> From: Paul Eggert <eggert at>
> For the 1996 EC / EU daylight savings changes, we need at least four
> rulesets, one each for Western, Middle, and Eastern Europe (since the
> change is UTC, not local time), and one for GB-Eire (since they will
> continue to use `GMT' and `BST' instead of `WET' and `WET DST').

Or implement my suggestion for a means of specifying a change at a UTC time.
The GB-Eire point about using GMT/BST is still valid, I can't see us changing.

> Perhaps the simplest thing is to use the existing ruleset names,
> using a comment to explain that as of 1996 the [WME]-Eur and GB-Eire
> rulesets use the EC rules.

Remember that there isn't anything magic about the EC rules in this seventh
Directive. They are just the same sort of EC rules as in the first to sixth
Directives. The only thing that changed was the agreement of a common end
date. The UK/Eire end date in previous Directives was just as much an EC
rule as any other.

In particular, all EC Directives have used a changeover time of 01:00 UTC,
right back to 1981. (At least, I am pretty sure they did, I don't have
the text for the 1st--5th, but the UK changed from its traditional 02:00 GMT
to 01:00 in 1981.)

> The biggest near-term problem in this area will be countries that
> currently differ from the EC rules only in that they switch at
> midnight UTC instead of the new standard 01:00.  It will be a hassle
> to find exactly when they change their rules.  We found out with
> Finland only because someone sent us a bug report.  Countries in this
> category currently include Albania, Bulgaria, Estonia, Greece, Latvia,
> Lithuania, Moldova, Poland, Romania, Turkey, and Ukraine.  Greece is
> the only country in this list that is required to change because it's
> in the EC, but most likely the rest will follow at some point.

Greece should have changed in 1981, it joined in the year the first Directive
took effect. I have no idea whether it actually did I'm afraid.

		Peter Ilieve		peter at

More information about the tz mailing list