Minor nit:

On Apr 21, 2010, at 04:57 , Richard Hainsworth wrote:
If a calendar system, eg., Chinese, Muslim and Jewish, defines days in the same way, eg., starting at midnight and incorporating leap seconds, for a time-zone, then the naming of the days is done by

The Jewish, Muslim, and Baha`i (and, technically at least, Christian) religious calendars use sunset as the start of the day. Only Western Christianity and Baha`i use the Gregorian corrections; Orthodox Christianity and the other aforementioned calendars are Julian. (Actually, the Baha`i calendar seems unclear on this; everything I can find suggests that it's still an open issue for them, but for the moment at least it's tied to the Gregorian calendar.)

The Jewish calendar doesn't recognize leap seconds; all time measurements are variable, with the day and night separately divided into 12 sha`ot composed of 60 dakot whose lengths vary as the length of the day/night changes throughout the year. Sunrise and sunset are usually defined as when the leading (resp. trailing) edge of the solar disk crosses the horizon. (Some communities use a variation where sunrise is considered to be "first light" and sunset the end of twilight; *usually* these are taken as the sun being 16° below the eastern horizon and 7.5° below the western. However, there's a lot of variation between communities, and very little of this is fixed; one opinion sets the start of "night" as 1 1/4 sha`ot *before* sunset!)

Islam has some other sticking points in this area: some communities define the start of a new month based on local time, while others base it on observations at Mecca, with the result that the calendar can vary by a day even in the same location. (This got some discussion in the context of Iraqi sectarian violence a few years back.) Additionally, while Jews use a fixed calculation of when the new moon occurs, Muslims still base the start of the month on direct observation of the new moon, so the calendar can again be off by a day.

All of which suggests that it may not actually be possible to come up with a mechanism suitable for representing all of these calendars.

brandon s. allbery [solaris,freebsd,perl,pugs,haskell] allb...@kf8nh.com
system administrator [openafs,heimdal,too many hats] allb...@ece.cmu.edu
electrical and computer engineering, carnegie mellon university    KF8NH

Attachment: PGP.sig
Description: This is a digitally signed message part

Reply via email to