RE: Question about legacy coding style

Thanks for all of the replies.

This type of thing always strikes me as being "programming archeology", in
that you had to be there to really know why something was done (what it is
doing is not too hard to discern, but the "why's" become dimmer over time).

There should probably be some project launched as a wiki to collect and
document these kinds of idioms. Somebody 10 years from now won't have any
idea why this was done, and those who remember, well, they might not be on
this list. Legacy COBOL is probably the same way (to pick another widely
used language with an enormous legacy code base).

Craig Pelkie

-----Original Message-----
From: rpg400-l-bounces@xxxxxxxxxxxx [mailto:rpg400-l-bounces@xxxxxxxxxxxx]
On Behalf Of Joe Pluta
Sent: Wednesday, March 23, 2011 9:09 PM
To: RPG programming on the IBM i / System i
Subject: Re: Question about legacy coding style

On 3/23/2011 10:52 PM, Joe Pluta wrote:

On 3/23/2011 10:38 PM, Booth Martin wrote:

Another possibility was that the programmer just wanted all of his
field definitions for each grouping of work fields together. I do
not understand why he would choose two lines though. A search for

"U1NU1"

would be a logical way for a programmer to have a common practice of
defining work fields.

U1NU1 turns into two lines of code when you convert RPG III to RPG/400
using CVTRPGSRC. RPG/400 only has room for one indicator per line.

Joe

Ugh. Got that confused. RPG/400 uses the original column sizes
(5-character opcodes, three indicators per line), while RPG IV is the newer
style (10-character opcode, one indicator). So the line above should have
read "convert RPG III (or RPG/400) to RPG IV" and "RPG IV only has room".
Gotta stop posting this late...

Joe