gEDA-dev: Another autosave crash....

[Ales Hvezda]

Hi Peter,

[snip]
>Indeed, not a fantastic idea by any means. I did wounder if the preview
>widget being visible might be an increased risk-factor in triggering the
>bug, as it did seem more common after I'd opened and closed the
>component selector a few times. (Noticed at the same time I discovered
>the crasher which Patrick identified as being due to print_command not
>initialised)

	Yes, I'm sure this added quite a bit to recent rash of crashes.
If people spend a little time testing the CVS version and report back
on their status, I will cut a new release of gEDA/gaf.  I don't want
this crashing to turn into the same problem like the gslice problem
(which was only a problem because it took me so long to fix the problem).
It is very important to init all toplevel variables in s_toplevel_new().


>
>
>Do we have any outstanding crashers which people are seeing on a regular
>basis? There are some listed on Sourceforge, but I don't know how many
>are now fixed.

	I have gone through and closed off a whole bunch that I think 
were due to the g_slice bug.

>
>The bug submitted today by Clunixchit (clunixchit) sounded like gslice
>thing from the backtrace, but I couldn't reproduce it by the steps he
>mentioned. (I wasn't running the release version, but CVS, so perhaps
>this matters).

	Yeah, that one was definately the g_slice bug. 

								-Ales