+1 - exactly my thoughts - all the convoluted mechanisms used - problems
that come from moving things in and out of development libraries -
having a single source repository - many other things were mentioned,
and all are just handled by an SCM.
As for knowing whether source matches - there are always the source
attributes of a program object or module object - those CAN be compared,
especially the create and change dates of source, compared to that
information that IS stored in every program or module object.
If objects have *SRCDBG info for OPM, then the source IS stored in the
Please, I understand there could be exceptions - so no need for anyone
to point that out, unless they have specific cases.
On 4/27/2014 9:57 AM, Paul Nelson wrote:
These are all good reasons to invest in a change management tool.
[mailto:midrange-l-bounces@xxxxxxxxxxxx] On Behalf Of Wilson Jonathan
Sent: Sunday, April 27, 2014 8:22 AM
To: Midrange Systems Technical Discussion
Subject: Re: how to increase field size without increasing the field size
On Sun, 2014-04-27 at 08:41 -0400, Steve Richter wrote:
On Sun, Apr 27, 2014 at 8:31 AM, Wilson Jonathan
I've had similar problems as sites where source was moved into user
On Sun, 2014-04-27 at 07:28 -0400, Steve Richter wrote:
In defense of my colleague who is reluctant to recompile all, there is
code that is 15 years old in the system. That predates him. No one
knows for sure if the source code on the system is the actual source
that was compiled to create the program. Would be nice if source was
always stored in the program object.
specific "development" libraries, compiled, and then the program and
source _moved_ back to the "live" system... at which point the link
between the program and source was lost from the DSPPGM command.
Only by enforcing the rule that modded/new programs _must_ move the
source back to "live" and then re-compile was the link retained.
The other way was not to "book out" source into lots of user dev
libraries, but to have one large "dev" mirroring the "live" except for
on going changes, which would then have the source pushed back to the
live (once dev work was complete), but the compilation was always done
from "dev" with the pgm objects then pushed back to live, so again the
link was retained. The "live" source files were the backup to the real
"live/dev" source in the dev library, not totally intuitive but worked
well in practise for a small shop (circa 5 people) as un-related but
impacted development could be coordinated but probably would have been
useless in larger shops.