[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Bug-gnubg] unstable gnubg

From: Petr Kadlec
Subject: Re: [Bug-gnubg] unstable gnubg
Date: Fri, 12 Nov 2004 13:12:11 +0100


> It's really hard to find this bug, since I can't recreate this when I'm
> running in the debugger. Jon had a theory about some events failing. I'll
> take a look at this later today.

Yesterday, I played an interesting match against gnubg and in gnubg crashed 
(with that "...cannot move, cannot move" message). I was really angry, then, I 
cooled a little bit down. :-)

After some experimenting, I am now able to deterministically cause the error. I 
am not sure whether it is the same bug (or at least a connected problem), but 
it has identical symptoms. So, here it goes: the problem happens 100% if I 
click on a dice in the moment I cannot move (I dance on bar). I tested it by 
starting a new match, then editing the position so that gnubg has a prime and 
the remaining pieces standing before my "anti-prime" of 6 single pieces waiting 
to be killed. Make it gnubg's turn, end edit mode, force gnubg to play. He will 
obviously throw out some pieces, and as soon as he finishes the move, prepare 
the mouse cursor to the position in the middle of the board, where the dices 
appear. Initially, I was just clicking wildly in hope of randomly hitting a 
dice. Then, I recognized that there is enough time to just wait for the dice to 
appear and then click on it. I am even able to click many times, so that the 
"cannot move" message is not present just twice, but e.g. 10 times. After 
clicking OK, the assertion failure occurs.

I was thinking about diving into the code, but 1. I did not have much time, 2. 
I do not have much understanding of the Gtk-related code.

So I just hope this could help.


reply via email to

[Prev in Thread] Current Thread [Next in Thread]