Red Hat Bugzilla – Bug 427046
Crashes on save
Last modified: 2008-01-03 17:10:49 EST
Description of problem:
If you try to save game (or load the resulting save) after finishing the 2nd
objective in the first level (?), the game crashes.
Version-Release number of selected component (if applicable):
Steps to Reproduce:
I just spend an hour trying to reproduce this without any luck. I assume that
with the first level you mean the level you get when you choose new game, and
with the second object you mean getting the info to exit through the caves?
Saving after that and restoring (either after quiting or directly) works fine.
Are you using a special arch like powerpc perhaps?
Anyways closing this as works for me for now. The only thing which could lead to
this actually getting fixed for you, would be to install
blobAndConquer-debuginfo, and then run blobAndConquer under gdb in windowed
mode, cause the crash and do a backtrace (bt command in gdb) after that.
No, no special arch, just plain i686 on an intel pentium 4 (3ghz), Nvidia
ElCheapo card and Nvidia driver. This is strange, as I have always been able to
reproduce it - but now I wasn't. I did a few things differently, tough:
- lower resolution (1024x768 instead of 1600x1200)
- Windowed mode (why?)
- Running directly instead of using wrapper
- Have installed debuginfo rpm
- Running through gdb
I can try to run it without windowed mode and with my old resolution. I can
also, if I manage to reproduce, send you the savegame (just loading it causes
crash). Where do I find those?
(In reply to comment #2)
> No, no special arch, just plain i686 on an intel pentium 4 (3ghz), Nvidia
> ElCheapo card and Nvidia driver. This is strange, as I have always been able to
> reproduce it - but now I wasn't. I did a few things differently, tough:
> - lower resolution (1024x768 instead of 1600x1200)
That might be the cause, I was using 800x600 (the default)
> - Windowed mode (why?)
Because if a fullscreen game segfaults while running under the debugger SDL's
segfault handler doesn't get called, and thus things effectively freeze (the
game is still fullscreen and still has the mouse and keyboard grabbed).
> - Running directly instead of using wrapper
Shouldn't make a difference
> - Have installed debuginfo rpm
Shouldn't make a difference
> - Running through gdb
Might make a difference, but I tried without gdb too.
> I can try to run it without windowed mode and with my old resolution. I can
> also, if I manage to reproduce, send you the savegame (just loading it causes
> crash). Where do I find those?
That might be an idea, just tar up and attach the contents of:
Created attachment 290611 [details]
Reproced it. Loading the 2nd savegame crashes the game for me (it just "quits to
desktop", no freeze).
I have no GL desktop etc. enabled. Just thought that i should mention it.
Thanks, neither safe game works on my 64 bit machine, I'll take a look when I'm
behind a 32 bit machine and / or find the time to install a 32 bit Fedora on
Perhaps change status then?
Strange that there should be a difference in savegames between 32- and 64- bit!
Is it some kind of mem dump?
I've done some further digging into this, I think you will find that you can
reproduce your crash when saving to slot 2 or 3 no matetr how far you are in the
game, the other slots should work fine.
You probably already played blobAndConquer when it was still at version 0.90,
and they have changed the savegame format. They (upstream) explicitly reserves
the right to do that until blobAndConquer hits version 1.0 .
I agree that crashing on this isn't pretty though. I'll try and see if I can fix
I've given this some more thoughts and I've come to the conclusion that I
currently have other priorities then writing code to automatically remove the
old uncompatible save games, so I'm closing this.
Please let me know if you still have problems after doing
"rm -fr ~/.parallelrealities/blobAndConquer"