GNOME Bugzilla – Bug 517347
Broken pipe to AI starting new game
Last modified: 2008-03-11 01:41:37 UTC
Version: 2.20.0.1 What were you doing when the application crashed? just surfing the net and changing games. i've run 5-6 games with GNU chess and so on... Distribution: Ubuntu 7.10 (gutsy) Gnome Release: 2.20.0 2007-09-17 (Ubuntu) BugBuddy Version: 2.18.1 System: Linux 2.6.22-14-generic #1 SMP Tue Feb 12 07:42:25 UTC 2008 i686 X Vendor: The X.Org Foundation X Vendor Release: 10300000 Selinux: No Accessibility: Disabled GTK+ Theme: Human Icon Theme: Human Memory status: size: 0 vsize: 0 resident: 0 share: 0 rss: 0 rss_rlim: 0 CPU usage: start_time: 0 rtime: 0 utime: 0 stime: 0 cutime:0 cstime: 0 timeout: 0 it_real_value: 0 frequency: 0 ----------- .xsession-errors --------------------- self.__autoload()
+ Trace 189672
self.addPGNGame(pgnGame, None)
newGame = self.addGame(gameProperties.name, gameProperties.white.name, w, gameProperties.black.name, b)
player = g.addAIPlayer(blackName, self.__aiProfiles[profile], level)
player = AIPlayer(self.application, name, profile, level, description)
ai.Player.__init__(self, name, profile, level)
os.write(fromManagerPipe[1], data)
-------------------------------------------------- Traceback (most recent call last): File "/usr/games/glchess", line 20, in <module> start_game() File "/usr/lib/python2.5/site-packages/glchess/glchess.py", line 4, in start_game app.start() File "/usr/lib/python2.5/site-packages/glchess/main.py", line 1386, in start self.__autoload()
The AI monitor process got an exception, I've made this process more resiliant to exceptions: http://svn.gnome.org/viewvc/gnome-games/trunk/glchess/src/lib/ai.py?r1=7140&r2=7447 This problem has been fixed in the development version. The fix will be available in the next major software release. Thank you for your bug report.
Make sure that you are confident that this change to the AI code doesn't introduce any serious crasher bugs, since there is a code freeze in a few days, and the release of 2.22.0 very soon after that. Some people still remember the bug where glChess' AI code would kill the X server...
The changes look more severe than they are... I've tested all the failure conditions I know so not expecting any problems. I really wanted to close that last critical chess bug before 2.22 :)
Nice. I get the "monitor died" if I choose resign from the menu. Is this a bug or leftover debug info?
Created attachment 106943 [details] [review] Stops AI incorrectly being reported as dead after quitted
*sigh*, actually it is a bug. See patch, requesting code-freeze break as this causes the game result to be human win due to AI death when a resign occurs.
Vincent Untz: "Since it's late, I'll give a second approval if two people tell me they tested the patch and that it works fine, without any other regression."
No regression here. Anyone else?
No regression. And the problem is gone here.
Jason has committed, closing. Cheers guys.