ACS Atlas

Paul Eggert eggert at twinsun.com
Mon Jun 12 13:34:40 UTC 2000


   From: "Jamie" <jamie at maxnet-usa.net>
   Date: Sun, 11 Jun 2000 22:09:23 -0400

   Doris Chase Doane has a book out on Time Changes through out the world.
    I just downloaded your timezone information from the ftp.btx.com site
   and noticed that it doesn't seem like you really have very much
   information.

The tz database concentrates on time changes since 1970, for two
reasons.  First, the tz code is based on the POSIX.1 computer
standard, and portable POSIX.1 time stamps cannot occur before 1970.
Second, it makes the database a lot smaller if we partition the world
only into locations whose time histories differ since 1970, and ignore
any extra partitioning due to earlier differences.

It'd be nice if someone extended the tz database backward in time to
all known time stamp histories, but that would be a lot of work and
nobody has done it (yet).

   I'm trying to create an atlas that provides the correct time zone
   based upon not just the date but the town and place.

If you find errors in the tz database, please send email to
<tz at elsie.nci.nih.gov>.  For more complete coverage of dates before
1970, I recommend Shanks's time zone history atlases.  They do not
always agree with Doane, but they seem to be popular among astrologers
these days.  Please see:

http://www.astro.com/atlas/index.html
http://astrocom.com/software/ACS_PC_Atlas.htm
http://astrocom.com/books/xrefa.htm#SHANKS

   there are frequently situations where the uncertainty about the time was
   actually the norm rather than the rule.

You're quite right.  For example, what was the time zone of Wise,
Virginia on October 18, 1927?  According to
<http://www.astrodatabank.com/NMScottGeorgeC.htm>, Doane and Shanks
disagree and in this case Doane's probably correct.  Or, to take a
modern example, what is the time zone of Pangnirntung, Nunavut,
Canada?  The province disagrees with the city, so city and commercial
offices operate in one time zone and provincial offices operate on
another.  (How will your database address problems like Pangnirtung's?)



More information about the tz mailing list