From: Guy Harris
Sent: Saturday, December 02, 2000 1:09 AM
> On Fri, Dec 01, 2000 at 10:41:47AM -0500, Gilbert Ramirez wrote:
> > Some g_assert() calls here and in other column-related functions
> > would not be a bad idea.
>
> Those calls would replace the SIGSEGV with a SIGABRT.
>
> If the goal is to have it exit "cleanly" in that situation, that
> wouldn't help.
>
> However, I don't think it *should* exit "cleanly" in that situation, as
> per my other mail; if "clp" is null, it either means that some data
> structure internal to {Ethereal, Tethereal, Cethereal, ...} is
> corrupted, or some routine is passing a bad value to
> "get_column_format()" for some other reason, and it should die messily,
> so that
>
> 1) it's clear that there's a bug and it needs to be fixed
>
> and
>
> 2) there's probably going to be a core dump, which could make it
> easier to track down the problem.
>
Yes there was a problem with the code; I have fixed it and I did use the
core dump to trace the problem. Thus I do agree with your assertions
and don't want any code changes.
Jeff Foster
jfoste@xxxxxxxxxxxx
