I used .rehash on an associative array, a couple times the compiler gave me:
Assertion failure: 'fd->type->next->equals(treturn)' on line 880 in file
'func.c'
I would have to just remove it for it to work, and other times it worked
fine. I guess it has to do with surrounding code.

When I threw an exception from a module's static constructor, main() was still entered. Should this be? If so, why?

Remaining objects destructors aren't being called on program exit, unless I explicitly delete them.

> When I threw an exception from a module's static constructor, main() was still entered. Should this be? If so, why?

Doesn't sound good

> Remaining objects destructors aren't being called on program exit, unless
I
> explicitly delete them.

Does this include "auto" qualified ones?

> > When I threw an exception from a module's static constructor, main() was still entered. Should this be? If so, why?
>
> Doesn't sound good

Actually, this isn't always the case.  I realized it was because it was my
WinMain() calling _moduleCtor(), so it does make sense.

>
> > Remaining objects destructors aren't being called on program exit,
unless
> I
> > explicitly delete them.
>
> Does this include "auto" qualified ones?
>

This, too, isn't always the case. It isn't calling the destructor when I have a global reference and assign a new object to it in another function, like main. "auto" works correctly.

"Vathix" <Vathix@kernel.net> wrote in message news:b7t6ed$23n$1@digitaldaemon.com...
> This, too, isn't always the case. It isn't calling the destructor when I have a global reference and assign a new object to it in another function, like main. "auto" works correctly.

The global reference is the problem. The gc still regards it as being pinned by the global reference.