Re: Debug getting lost?

David Gibbs wrote:

On 12/3/2009 12:45 PM, Mark S. Waterbury wrote:

the breakpoint would be set in the old copy of
the *PGM that now resides in QRPLOBJ, renamed with a "Q" followed by
a hexadecimal timestamp. :-o

I set added a program to SEP debug, debugged it, recompiled it, refreshed it in the SEP view, and tried to debug it again.

But the 2nd time I try to debug it no breakpoints hit. This is even after signing off and back on in the green screen (where the program is running).

After refreshing the SEP debug entry and signing off and back on, the program should no longer be pointing to a QRPLOBJ version.

I don't see where there has been anything added to this thread for a week. What I want to comment on is that very last bit -- "the program should no longer be pointing to a QRPLOBJ version."

Why not? As far as I know, that QRPLOBJ version will exist unchanged until IPL or some manual deletion of the QRPLOBJ copy. What exactly should no longer be 'pointing' to it?

That is, if a "SEP" (whatever those actually are) is directly associated with the *PGM object that was moved to QRPLOBJ, then it will remain associated with it regardless of how many new copies result from recompiles. If a "SEP" is actually some bit of code inside of a *PGM object that gets activated during debug, it will still be there after the library changes to QRPLOBJ. It won't magically move into some different *PGM object simply because it has the same name. (Will it? Should it?)

The recompiles shouldn't affect that kind of SEP at all.

The problem seems to be that few (or none?) of us know what a "SEP" really is so we can't predict or understand how they're supposed to work.

Or am I misunderstanding the issue?

Tom Liotta