Re: Assertion in symbol table due to Revision 8881

[Daniel J Sebald]

John W. Eaton wrote:
> On 29-Jun-2009, Daniel J Sebald wrote:
>
> | The octave-core dump ends up being only 11 bytes long.
>
> The octave-core file is just an octave save file that contains the
> variables in the top-level workspace.  It is intended to help you
> avoid losing all your work if Octave crashes.  But it seems that it is
> a common misconception that this file can somehow be used for
> debugging.

Well, that was my understanding of octave-core.  I guess that misconception 
came about because I think you used to request the octave-core from time to 
time when someone reported a bug.  (Not that you ever said there was debug info 
in octave-core.)

 Maybe simply changing the name to be something like
> octave-crash-workspace-PID would help avoid the confusion?  If there
> is agreement, I can make this change.

I'm fine with octave-core; prefer it actually.  The reason is that files with 
"core" in them are often tossed as inconsequential unless pertaining to the 
context at hand.  I.e., to free space, people often recursively remove core dump files or 
leave them out of archives.

Dan