On Tue, 12 Jan 1999 14:38:15 -0500, Glenn Cole
<[log in to unmask]> wrote:

>Clive Pottinger writes:
>
>> NOTE: The number returned representing the Lotus 1-2-3 date is
>> inaccurate for dates before Mar 1, 1900.  This is because Lotus 1-2-3
>> has a bug - it incorrectly treats 1900 as a leap year.
>
>Incidentally, this touches on why I think we got "lucky" with y2k situation.
>
>From Clive's statement, it looks like 1-2-3 takes as its leap year rule
>"any year that is evenly divisible by 4."  Of course, that's only one
>piece of a 3-piece rule.  It is our good fortune that this piece also
>points to 2000 as being a leap year.
>
>If technology had advanced "a bit faster" and we were preparing for 1900,
>there would be more software failure.
>
>--Glenn
>
>.......................................................................
>
>Item Subject: cc:Mail Text
>

Hi, Glenn! Every time this comes up, it reminds me of a database
program (SynFile+) I worked with years ago on an Atari 800. I was
writing some db utilities and was puzzled when I discovered that the
db's date fields reckoned day 0 to be 3/1/1900 and the max day of
65535 was sometime in 2079. Later I realized it was so the program
didn't have to waste precious code space (only 48Kb of memory) on the
century exceptions of the leap year rule either in 1900 or 2100.