RE: Why RPG IV ....

["Jon Paris" <Jon.Paris@xxxxxxxxxxxxxx>]

 >> However, what about his initial question of who cares about what century
when going from 0001-01-01 to 01/01/01?

But there are already rules in place concerning how to interpret century.
01/01/01 in *MDY is Jan 1st 2001.

You can't suddenly change the interpretation much as we might dislike it -
0001-01-01 is the low value for a 4 digit century. 01/01/40 is lowval for a
*MDY.  So if you were to change to 01/01/01 on a display file output, then
if that is an I/O field it would cause chaos because even with no user input
it would change from lowval to a 2001 date!  I suspect there would be a lot
more complaints about that!  You can't always change 01/01/01 to lowval
either since it would preclude Jan 1st 2001 ever being used as an input
date.

Much as we hate the current system we're pretty much stuck with it.  There
was a suggestion some time back that an option be added to the F spec that
said "leave numbers and dates alone" - but nobody voted for it.  The only
other alternative is to have yet another compiler directive, and it still
wouldn't become the default.

If we'd had this debate ten years ago - or even at V4R2(?) when Rochester
added dates to display files ....


Jon Paris
www.Partner400.com