From: Daniel Martin <martin@...>
Date: 2006-08-19T03:02:43+09:00
Subject: Re: Microsoft Timestamp (Active Directory)

"Berger, Daniel" <Daniel.Berger@qwest.com> writes:

> Leave it to Microsoft to come up with something this convoluted:

At least their time 0 is close to the beginning of a century, the
beginning of a year, and the beginning of the week. (if you begin
weeks on a Monday) They could have chosen to fill their 64-bit time
value with 100ns offsets from 00:00 17-NOV-1858, which isn't even at
the beginning of anybody's week.  (It's a Wednesday)

Microsoft even seems to have gone with a pre-existing external
standard in this case.

Parenthetical note:

The date in 1858 is VMS's time zero - in this case it was also a
pre-existing standard that was followed, based on what astronomers
using computers to track Sputnik used for their time values - see
http://www3.sympatico.ca/n.rieck/docs/calendar_time_y2k_etc.html#nov-17-1858

See also the bottom of rfc868:
http://www.cse.ohio-state.edu/cgi-bin/rfc/rfc868.html

I want to know what systems had time clocks starting 1 Jan 1976; I
assume the 1 May 1983 date is related to the publication date of the
RFC.