I've been working on a custom resource editor using PG so I'm experiencing
my share of crashes and screwy performance as I wend my way through the
intricacies of working with resource files; and that's ok, I expect it.

Anyway... I was experiencing some strange crashes that didn't seem to be
fitting any pattern except that they were memory addressing errors. Some
times it would occur, then it wouldn't, then it would... Using the FB
debugger "made it go away" but commenting out sections of code wouldn't
(necessarily). Well, while I was trying to figure out what was going on a
T-storm wandered through and blasted the power for a second or 2 and after
I re-booted I ran the Norton Disk Doctor to check the condition of my disks
and, fortunately, everything checked out ok, with the usual exceptions of
bundle bits, dates, etc., but the Doc reported that the PG project resource
file was damaged and should be replaced from a backup.

I checked it again and got the same report. Opening the file in ResEdit
went without a hitch as did opening it in PG but I recreated it and, you
guessed it, the strange crashes disappeared!

As I mentioned, I expect to trash some resource files while I'm doing this
development and I'm not really surprised that the trashed file was causing
the screwy aborts but I can't say as I expected anything this tricky. How
come The Doctor detected a problem (I'm _really_ glad he did;-) that
neither PG nor _ResEdit_ could and what can be the nature of such a subtle
clobber?

Any clues?


Charlie Dickman
charlied@...