Wirt asks:
> In that, UNIX blows the doors off of QueryCalc. I limited QC's calendar
> to only the years 1583 to 9999. 1583 was the first complete year of the
> modern Gregorian calendar. As the author of a report writer, I didn't
...
> My question is: what does UNIX do about the missing days in October,
> 1582? And if UNIX really does handle October correctly, why would it


The basic answer is:

   *no* Julian/Gregorian calendar makes a lot of sense prior to 1752-09-14.
   ...and possibly even later than that.   ...  forget about 1582!

Why?

   Many western countries finally adopted the Gregorian Calendar in
   September, 1752 ... dropping  almost two weeks:

      cal 9 1752

         September 1752
       S  M Tu  W Th  F  S
             1  2 14 15 16
      17 18 19 20 21 22 23
      24 25 26 27 28 29 30

I seem to recall that a few countries (including Russia?) waited even
longer to adopt the Gregorian calendar ... yes, found it on the net:

   From: http://genealogy.org/~scottlee/cal-overview.html  ...

   The Gregorian calendar was not instituted until October 15, 1582
   (or October 5, 1582 in the Julian calendar). Some countries did not
   accept it until much later. For example, Britain converted in 1752,
   The USSR in 1918 and Greece in 1923. Most European countries used the
   Julian calendar prior to the Gregorian.

So the bottom line is that unless you define the adoption point of the
Gregorian (and earlier) calendars for your particular pre-1923 calendar date,
you have little possibility of being "correct".

The Unix "cal" command appears to follow the British switchover date,
of 1752-09-14 (Gregorian)

--
Stan Sieler                                          [log in to unmask]
                                     http://www.allegro.com/sieler.html