In message <Pine.SGI.4.10.10010060800120.2159-100000@maanantai.hit.fi>, Petri K
oistinen writes:
>Oh, I didn't know that. I am not a expert, as quite unexperienced user, my
>opinion is that no program should give core dump, if there has not happen
>a malfunction.

Well, core dumps can be useful things. The knowledge that you can
send a signal from the keyboard that can cause a program to core dump
can certainly be useful.

> After all is that so hard to fix?

It's certainly not hard behavior to change. The question is whether
and why we would want to change it.

It's certainly the case that this has been the existing behavior
for a long time, and many people are used to it, and others find it
useful. Those are good reasons to keep it. I'm not sure what reasons
there are to take it away.

Furthermore, it's relatively portable behavior -- even Linux does it ;-)

>Anyway I was just trying to help, if you consider this normal
>operation then it's OK, I still personaly consider it bit odd
>feature.

It's potentially "odd" if you're not expecting it. But so is the behavior
of many of the other special characters that are enumerated if you run
"stty -a".

>Should programs just stop/quit if they get SIGQUIT?

Well, the default behavior when receiving SIGQUIT is to print Quit
and core dump. This is the default behavior for a lot of signals,
and is documented.

--jhawk