design-time access violations no longer showing a MadExcept stack trace

delphi package - automated exception handling
Post Reply
zbeleh
Posts: 15
Joined: Tue Jan 19, 2021 7:14 pm

design-time access violations no longer showing a MadExcept stack trace

Post by zbeleh »

hello,

at some point in time, i recall clearly dropping a design-time component on a form, which caused an access violation.
that one time, a MadExcept bugreport showed up with full stack trace information.

but for some reason, this is no longer showing, and i only get the standard Delphi "access violation" dialog, which is not very helpful.

in "Tools > MadExcept", the IDE exception catching is checked (the IDE design-time package is installed).
maybe there is some file missing from the PATH environment variable or so?

Any ideas?
Thanks!
madshi
Site Admin
Posts: 10754
Joined: Sun Mar 21, 2004 5:25 pm

Re: design-time access violations no longer showing a MadExcept stack trace

Post by madshi »

I've removed this in newer Delphi versions because it caused the IDE to become unstable. I'm not 100% sure why, I think it's likely an IDE problem, but what can I do. You can force it back by manually adding the madExceptIde_.bpl to the list of loaded packages (component -> install packages). Maybe you're lucky and the IDE stays stable that way.
zbeleh
Posts: 15
Joined: Tue Jan 19, 2021 7:14 pm

Re: design-time access violations no longer showing a MadExcept stack trace

Post by zbeleh »

sorry for the late answer, i just checked: the aforementioned BPL has been in my installed packages for a long time (manually installed).
yet I am unable to get design-time access-violations stack trace.

is there anything else I should be checking?
thanks!
madshi
Site Admin
Posts: 10754
Joined: Sun Mar 21, 2004 5:25 pm

Re: design-time access violations no longer showing a MadExcept stack trace

Post by madshi »

I'm not sure why it isn't working for you. My best guess is this: Generally, madExcept only cares about unhandled exceptions. So if maybe the Delphi IDE has a try..except block somewhere, which completely handles the exception (without re-raising it), then madExcept won't become active. Maybe that is what's happening here? This could potentially result in some IDE exceptions being caught and shown by madExcept and others not, depending on whether the IDE has a try..except block handling any specific exception or not.
Post Reply