My confusion with Asia/Aqtau

[Paul Eggert]

"Olson, Arthur David (NIH/NCI)" <olsona at dc37a.nci.nih.gov> writes:

> I don't know *really* happened in Asia/Aqtau in 1995;

Nor do I, alas.

> if there was indeed just one change to clocks, the "1995 Sep lastSun" line
> needs tweaking.

Since I invented the 1995 transition in the tz data, I suppose I
should be the one to clean it up....

We don't know when in the year the transition occurred.  I'm not even
100% sure the transition exists; it's derived only from the IATA data,
which are fallible.  But, on further thought, it would have made far
more sense to make the change in spring than in fall.  So I'm inclined
to make the following fix, which I'll incorporate in my next proposed
patch.

@@ -926,7 +901,7 @@ Zone	Asia/Aqtau	3:21:04	-	LMT	1924 May  
 			6:00	-	SHET	1982 Apr  1
 			5:00 RussiaAsia	SHE%sT	1991
 			5:00	-	SHET	1991 Dec 16 # independence
-			5:00 RussiaAsia	AQT%sT	1995 Sep lastSun # Aqtau Time
+			5:00 RussiaAsia	AQT%sT	1995 Mar lastSun 2:00 # Aqtau Time
 			4:00 RussiaAsia	AQT%sT
 # West Kazakhstan
 Zone	Asia/Oral	3:25:24	-	LMT	1924 May  2 # or Ural'sk



PS to Dave Rolsky: Thanks for reporting the bug.