On 3/27/06, Michael van der Gulik <squeakml at gulik.co.nz> wrote:

> This is a rather philosophical question: is it a bug if the VM crashes,
> showing either a General Protection Fault, Segmentation Fault and core
> dump, or whatever the local operating system does?

It's _somebody_'s bug. Unless it's your desired outcome, of course. :-)

> So the question, again, is: should the VM *never* crash and protect
> against very curious, stupid programmers like myself, or should it work
> fine until the programmer becomes curious and stupid?

A sufficiently curious programmer is capable of crashing any system;
stupidity is not a requirement. :-)

Still, in some of the places where the VM crashes today, it may have a
way to recover tomorrow. It should IMHO be sufficiently robust that it
doesn't crash until you've done the software equivalent of dropping a
wrench in the gears. In other words, when you crash it, you should
know (or easily be able to determine) why it's your own fault. If
Squeak isn't this stable today, it's pretty close.

If you find that the VM crashes when you think it should recover,
maybe you can cook up a patch. Or, at least, file a bug report on
Mantis. We'll never see a Squeak that can't crash, but it should be
stable enough for all normal uses.

Cheers!

--Tom Phoenix