gnugo-devel
[Top][All Lists]
Advanced

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

Re: [gnugo-devel] huh, =?iso-8859-1?q?wtf=3F


From: Paul Pogonyshev
Subject: Re: [gnugo-devel] huh, =?iso-8859-1?q?wtf=3F
Date: Sat, 21 Jun 2003 18:31:20 -0400

?=
Date: Sun, 22 Jun 2003 01:46:06 -0400
User-Agent: KMail/1.5.9
References: <address@hidden>
In-Reply-To: <address@hidden>
X-KMail-Link-Message: 1101
X-KMail-Link-Type: reply
MIME-Version: 1.0
Content-Disposition: inline
Content-Type: text/plain;
  charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
Message-Id: <address@hidden>
Status: RO
X-Status: Q
X-KMail-EncryptionState:  
X-KMail-SignatureState:  
X-KMail-MDN-Sent:  

Gunnar wrote:
> Paul wrote:
> > can anyone explain me this piece of code (interface/play_gtp.c,
> > around lines 2690, finish_and_score()):
> >
> >   do {
> >     move_val = genmove_conservative(&i, &j, next);
> >     play_move(POS(i, j), next);
> >     if (move_val >= 0) {
> >     pass = 0;
> >     moves++;
> >     }
> >     else
> >     pass++;
> >
> >     next = OTHER_COLOR(next);
> >   } while (pass < 2 && moves < board_size * board_size);
> >
> > what does `moves' serve for?
>
> Isn't that obvious? It breaks the loop in case some bug should cause
> GNU Go to play on forever, e.g. getting stuck in a triple ko. Limiting
> the number of moves to the number of vertices on the board is of
> course arbitrary but large enough not to happen in practice (in
> absence of bugs) and small enough to guarantee ending the scoring in
> reasonable time.

ok, i was to quick to decide on this :) but try to load the attached game
and enter `final_score'. gnu go enters a (seemingly) infinite loop.
something is definetely wrong.

Paul




reply via email to

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