Well, the first half of my code ran just fine yesterday (as far as we can tell.)
The second part abended. Seems the black box within the black box didn't like *something* in the data -- so we put some unnecessary and unwanted fields into the file we were trying to load, and it loaded just fine... (data problem? Not really -- the values there were correct as-was, so we need to find a way to trick the black box!)
Then we started to look at certain people. One specific value we wanted to find on about 30 folks we only found on 10. So the code (within the black box) is for some reason not-working on some non-zero (but not-100) percent of the rows.
Quite annoying. If it were in a section that I was just manually shoving out there, I'd have a hope of being able to stick diagnostics in, but as it is, I don't know how I'll figure out how to debug it. (I will find a way, though, stubborn cuss that I am!)
It's a good thing we're testing it in the stress environment, though, since there are far fewer changes than would have been the case in the Development environment that will need to be made manually to "make it right" ... and I have a week or so until I need to set it loose "for real" in Dev. (The only good part is that the ... problem seems to be consistent in the 2 environments, so I have a CHANCE at being able to test it.)