[tz] Proposed time zone package changes (Bahia Brazil, Pridnestrovie, Ukraine)

[Alan Barrett]

On Thu, 20 Oct 2011, Robert Elz wrote:
>Here are proposed changes to the time zone package.   The change
>to Bahia Brazil has already occurred, and the others are due to
>occur within 10 days (the weekend after next) so updates have become
>urgent (yet again)...

[...]

>Please test these changes (particularly those of you using 64 bit time_t
>systems).

I have tested these changes on a NetBSD system with 64-bit time_t.  I
find the following differences in the output from "zdump -v ${ZONENAME}".

America/Bahia gains two DST transitions per year, from now until further
notice:

  America/Bahia  Sun Feb 16 02:00:00 2003 UTC = Sat Feb 15 23:00:00 2003 BRT isdst=0
+America/Bahia  Sun Oct 16 02:59:59 2011 UTC = Sat Oct 15 23:59:59 2011 BRT isdst=0
+America/Bahia  Sun Oct 16 03:00:00 2011 UTC = Sun Oct 16 01:00:00 2011 BRST isdst=1
+America/Bahia  Sun Feb 26 01:59:59 2012 UTC = Sat Feb 25 23:59:59 2012 BRST isdst=1
+America/Bahia  Sun Feb 26 02:00:00 2012 UTC = Sat Feb 25 23:00:00 2012 BRT isdst=0
  [...]

The Europe/Kiev clock change that already occurred on 27 March
2011 is retroactively re-labeled as a transition into DST, instead
of a change in the UTC offset used for standard time.  The
displayed local time on 27 March 2011 remains the same, but the
zone name and isdst flag were "FET isdst=0" in tzdata2011l, and
change to "EEST isdst=1" in the new data:

  Europe/Kiev  Sun Mar 27 00:59:59 2011 UTC = Sun Mar 27 02:59:59 2011 EET isdst=0
-Europe/Kiev  Sun Mar 27 01:00:00 2011 UTC = Sun Mar 27 04:00:00 2011 FET isdst=0
+Europe/Kiev  Sun Mar 27 01:00:00 2011 UTC = Sun Mar 27 04:00:00 2011 EEST isdst=1

Europe/Kiev gets two DST transitions per year from now until further
notice:

+Europe/Kiev  Sun Oct 30 00:59:59 2011 UTC = Sun Oct 30 03:59:59 2011 EEST isdst=1
+Europe/Kiev  Sun Oct 30 01:00:00 2011 UTC = Sun Oct 30 03:00:00 2011 EET isdst=0
+Europe/Kiev  Sun Mar 25 00:59:59 2012 UTC = Sun Mar 25 02:59:59 2012 EET isdst=0
+Europe/Kiev  Sun Mar 25 01:00:00 2012 UTC = Sun Mar 25 04:00:00 2012 EEST isdst=1
  [...]

In tzdata2011l, Europe/Tiraspol was a link to Europe/Chisinau.  In the
new data, Europe/Tiraspol is a separate zone.  This change affects times
up to and including 31 December 1879, because the longitute used for LMT
has changed from that of Chisinau to that of Tiraspol:

-Europe/Tiraspol  Wed Dec 31 22:04:39 1879 UTC = Wed Dec 31 23:59:59 1879 LMT isdst=0
-Europe/Tiraspol  Wed Dec 31 22:04:40 1879 UTC = Wed Dec 31 23:59:40 1879 CMT isdst=0
+Europe/Tiraspol  Wed Dec 31 22:01:27 1879 UTC = Wed Dec 31 23:59:59 1879 LMT isdst=0
+Europe/Tiraspol  Wed Dec 31 22:01:28 1879 UTC = Wed Dec 31 23:56:28 1879 CMT isdst=0

The Europe/Tiraspol clock change that already occurred on 27 March
2011 is retroactively re-labeled as a change in the UTC offset for
standard time, instead of a transition into DST.  The displayed
local time on 27 March 2011 remains the same, but the zone name
and isdst flag were "EEST isdst=1" in tzdata2011l, and change to
"FET isdst=0" in the new data.  Also, all future DST transitions
for Europe/Tiraspol are removed in the new data:

  Europe/Tiraspol  Sun Mar 27 00:59:59 2011 UTC = Sun Mar 27 02:59:59 2011 EET isdst=0
-Europe/Tiraspol  Sun Mar 27 01:00:00 2011 UTC = Sun Mar 27 04:00:00 2011 EEST isdst=1
+Europe/Tiraspol  Sun Mar 27 01:00:00 2011 UTC = Sun Mar 27 04:00:00 2011 FET isdst=0
-Europe/Tiraspol  Sun Oct 30 00:59:59 2011 UTC = Sun Oct 30 03:59:59 2011 EEST isdst=1
-Europe/Tiraspol  Sun Oct 30 01:00:00 2011 UTC = Sun Oct 30 03:00:00 2011 EET isdst=0
-Europe/Tiraspol  Sun Mar 25 00:59:59 2012 UTC = Sun Mar 25 02:59:59 2012 EET isdst=0
  [...]

I also noticed a mistake in a comment in the "europe" input file:

+# The substance of this change (reinstatement of the Tiraspol entry)
+# is from a patch from Petr Machata (2011-11-17)

That date should be 2011-10-17.


--apb (Alan Barrett)