TZ database content

[John A. Halloran]

At 03:07 PM 02/12/01 +1100, Alex LIVINGSTON <alex at agsm.edu.au> wrote:
>At 09:50 -0800 2001-02-09, John A. Halloran wrote:
>>Pasted in here is a complete list of time zones and abbreviations.
>
>Certainly not complete, by far. Where's Nepal, Lord Howe Island, the 
>Chatham Islands, Burma (a.k.a. Myanmar), Cocos Islands, Norfolk 
>Island, Marquesas Islands, Tonga, Fiji, Afganistan, ...?
>
>>     TimeZones(41) = "South Australian Summer Time - SDT (-11:30)"
>>     TimeZones(42) = "South Australian Time - SAT (-10:30)"
>
>Where did these come from?

Alex Livingston,

Thank you for drawing my attention to these special time zone locations.
My list has a couple of holes, but not as many as you suggest.

You mention Lord Howe Island and then ask where South Australian Time with
a standard of -10:30 comes from.  Lord Howe is the only Australian location
that observes South Australian Time.  The daylight saving offset appears to
have changed however from one hour to half an hour in 1985.  There is an
Australian time zone for this offset, but the abbreviation would not
correlate to the correct standard time zone.

Myanmar/Burma and the Cocos Islands, both standardized at -6:30, are in the
list as North Sumatra Time.

Norfolk Island at -11:30 is standardized to what the list calls New Zealand
Time.

The Marquesas Islands in French Polynesia at -9:30 are standardized to what
the list calls Australian Central Standard Time.

The list certainly does not have the -12:20 standard of Tonga before 1968,
but it does have its -13:00 standard since then.  The list calls that
Chukot Time.

Fiji standardized at -12:00 uses the zone that the list calls New Zealand
Standard Time.

Afghanistan is standardized at -4:30, which is in the list as Iran Daylight
Time.  Afghanistan could have its own zone.

Nepal at -5:45 since 1986 is not in the list.

I was not aware of the Chatham Islands east of New Zealand and east of the
International Date Line at -12:45 and -13:45 in summer.

Since politicians create time zones, there could always be new exceptions
to any attempt to freeze the world time zone situation in a comprehensive
list.  And I do agree with some commentators that trying to force world
time zones into tidy boxes may come from an American mindset conditioned by
American time zone practices.

If the politicians can change the daylight offset from 1 hour to 1/2 hour,
it would take awhile for a new abbreviation to percolate to the surface
that would reference that daylight time zone. But for any case where a
program cannot match the time zone amount to a known time zone name, the
program will just call it Local Time and will proceed merrily on its way.

The numeric amount is primary.  Abbreviations and names must be stored with
their numeric amounts, and abbreviations tested mainly to identify the
desired zone name when more than one named zone has the specified numeric
amount.  Zone names are important for the human interface, to be matched
and linked to when possible.  Don't discount the importance of the human
interface.  In recognition of it, the trend for data exchange now consists
of human readable, meaningful XML element tags, as opposed to obscure fixed
field record formats that require a computer to make sense of them.

Regards,

-------------------------------------------
John Halloran, Halloran Software
P.O. Box 75713, Los Angeles, CA 90075 U.S.A.
http://www.halloran.com/
e-mail: seagoat at primenet.com