FW: Does zic currently compile time zone abbreviations correctly?

Olson, Arthur David (NIH/NCI) [E] olsona at dc37a.nci.nih.gov
Fri Jul 27 15:20:33 UTC 2007

I'm forwarding this message from John B. Morrow, who is not on the time zone mailing list.

Those of you who are on the time zone mailing list should direct replies appropriately.


From: John B. Morrow [mailto:jmorrow at sendwordnow.com] 
Sent: Friday, July 27, 2007 2:29 AM
To: tz at lecserver.nci.nih.gov
Subject: Does zic currently compile time zone abbreviations correctly?

In the process of extracting time zone information from the zoneinfo database, I noticed that the abbreviation pointer break for quite a few time zones including Europe/London, Asia/Dubai, and quite a few others.  The problem seems to be in the compiled time zone files so I spent some time looking at zic.  What seems to be happening is that zic uses some criteria to determine of the abbreviation (and other info for types) is used and won't write it if it thinks it isn't needed.  The problem is that it doesn't adjust the indexes (tt_abbrind) that reference those abbreviations, so they'll point to bad data or empty strings.  My quick (and admittedly heavy handed fix) was to force zic to write all of the abbreviations, whether they are necessary or not.  If this really is a problem, it should probably be fixed at some point.  Please note that the data in our fairly recently updated Linux server has abbreviations that don't seem to work.
If I'm missing something or there is some other way more correct way to reference the abbreviations, please let me know.  I'm using tt_abbrind as per the tzfile man page and the tzcode2007f.tar.gz sources. 
John Morrow

More information about the tz mailing list