hoichess died

[GoogleCodeExporter]

see the attached log

Original issue reported on code.google.com by gbtami on 23 Oct 2007 at 4:35

[GoogleCodeExporter]

khmm

Original comment by gbtami on 23 Oct 2007 at 4:37

Attachments:

• 2007-10-23_18-25-30.log
• hoichess.png

[GoogleCodeExporter]

Can you judge, whether it might have been killed for doing an illegal move?
The newest revisions should always show the exitcode of killed engines.
The version, with which this was reported, doesn't seam to do so.

Original comment by lobais on 2 Dec 2007 at 9:39

• Changed state: Need-Info

[GoogleCodeExporter]

Original comment by gbtami on 2 Dec 2007 at 10:35

Attachments:

• 2007-12-02_11-15-02.log
• hoichess.png

[GoogleCodeExporter]

Ok, so we can conclude it isn't killed by PyChess due to an illegal move.

Revision 788 no longer ignores sigchild, so we should be able to get the 
exitcodes.

To report this kind of errors to the engine-developers, it would be interesting 
to
use functions such as:

WCOREDUMP(status)
    Returns True if a core dump was generated for the process, otherwise it returns
False. Availability: Macintosh, Unix. New in version 2.3. 
(from http://docs.python.org/lib/os-process.html)

It doesn't seam to return the actual core dump though.

Original comment by lobais on 2 Dec 2007 at 11:37

[GoogleCodeExporter]

The tail of log file with rev 788:

15:04:41 Hoichess Log: result 1-0 {?}
15:04:41 Hoichess Log: quit
15:04:41 Hoichess Debug: Exitcode 0 Success
15:04:47 Hoichess Debug: /usr/games/hoichess
15:04:47 Hoichess Debug: Black
15:04:47 Hoichess Log: xboard
15:04:47 Hoichess Log: protover 2
15:04:47 Hoichess Log: nopost
15:04:47 Hoichess Log: hard
15:04:47 Hoichess Log: sd 9
15:04:47 Hoichess Log: level 0 5 0
15:04:50 Hoichess Log: time 300
15:04:50 Hoichess Log: otim 300
15:04:50 Hoichess Log: e2e4
15:04:50 Hoichess Log: quit
15:04:50 Hoichess Debug: Exitcode 0 Success

Original comment by gbtami on 2 Dec 2007 at 2:07

[GoogleCodeExporter]

Ok.
791 has been upgraded with massive debugging for stuff like this.
I'm eager to see how it goes.

Original comment by lobais on 9 Dec 2007 at 4:40

[GoogleCodeExporter]

On game ends i get this:

20:02:53 Default Debug: Ending a game with status 5 for reason 20
File "./pychess", line 41, in <module>
    pychess.Main.run(sys.argv[1:])
  File "/home/tamas/pychess-philidor0.8beta2/lib/pychess/Main.py", line 477, in run
    gtk.main()
  File "/home/tamas/pychess-philidor0.8beta2/lib/pychess/widgets/BoardControl.py",
line 231, in actionActivate
    self.emit("action", RESIGNATION, None)
  File "/home/tamas/pychess-philidor0.8beta2/lib/pychess/Players/Human.py", line 71,
in <lambda>
    self.board.connect("action", lambda b,ac,pa: self.emit_action(ac,pa))
  File "/home/tamas/pychess-philidor0.8beta2/lib/pychess/Players/Human.py", line 101,
in emit_action
    self.emit("offer", Offer(action, param))
  File "/home/tamas/pychess-philidor0.8beta2/lib/pychess/Utils/GameModel.py", line
140, in offerRecieved
    self.end(BLACKWON, WON_RESIGN)
  File "/home/tamas/pychess-philidor0.8beta2/lib/pychess/Utils/GameModel.py", line
401, in end
    "".join(traceback.format_list(traceback.extract_stack())).strip()))

Original comment by gbtami on 9 Dec 2007 at 7:03

[GoogleCodeExporter]

Lol :S
I 'spose you didn't click that resign menu item yourself.
Perhaps a few prints on the way from keyboard event to emit_action would help
pinpoint the source.

Original comment by lobais on 9 Dec 2007 at 7:48

[GoogleCodeExporter]

Nope.I myself was who clicked the resign menu.

Original comment by gbtami on 10 Dec 2007 at 9:22

[GoogleCodeExporter]

Huh. Then it can't be seen as a problem that the game ended?

Perhaps the problem is the output format, which copies the standard python 
uncaught
error format.

Didn't hoichess die you say?

Original comment by lobais on 10 Dec 2007 at 3:24

[GoogleCodeExporter]

There doesn't seam to be a problem her anymore.
Please reopen if I'm wrong.

Original comment by lobais on 21 Dec 2007 at 1:26

• Changed state: Invalid