[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Bug-gnubg] Tutor Cube Warning bug and Export Analysis Bug
From: |
Ned Cross |
Subject: |
Re: [Bug-gnubg] Tutor Cube Warning bug and Export Analysis Bug |
Date: |
Wed, 2 Oct 2002 10:02:44 -0700 |
I don't know if I misrepresented the problem when I referred to "redoubling
to 4". I think in some areas of the GNUBG program the term "redouble" is
synonymous with the term "beaver" ( at least that's what Help/Commands
says).
I don't know if it makes a difference, but I was actually not beavering
here. The cube had been on 2 for some time and I was doubling to 4.
Also I have seen GNUBG warn of bad cubes to 4 when not good enough to
double, so the cube tutor seems to work to some extent.
Ned
----- Original Message -----
From: "Jim Segrave" <address@hidden>
To: "Ned Cross" <address@hidden>
Cc: "GNUBackgammon bug reporting" <address@hidden>
Sent: Wednesday, October 02, 2002 2:48 AM
Subject: Re: [Bug-gnubg] Tutor Cube Warning bug and Export Analysis Bug
> On Tue 01 Oct 2002 (12:24 -0700), Ned Cross wrote:
> > OS: Win ME
> > GNUBG: Version 0.12 Build 020930
> > Bug: 1) Tutor mode fails to give warning flag for re-cube when TG
> >
> >
> > 1) I was playing out a midgame position from a previous match, using
tutor mode with the following settings:
> > Cube decisions - on
> > Checker play - on
> > Tutor Decisions - Same as Evaluation
> > Evaluation settings - (Checker Play and Cube settings Identical)
2-ply, 6 candidates, tol. 0.110, Reduced 2-ply, cubeful checker eval.
> > Warning Level - Doubtful
> >
> > **BUG** Redoubling to 4 in the posted position provoked no warning flag
from the tutor. GNU passed, and post-game analysis flagged the recube as
"bad" - an error of -0.130.
>
> Hmm - someone has commented this code out. From play.c:
>
> #if 0
>
> removed until gnubg/GoodDouble handles beavers and raccoons
> correctly.
>
> if ( fTutor && fTutorCube && !GiveAdvice( GoodDouble( TRUE, pmr )
> ))
> return;
> #endif
>
> Joern took this out in rev 1.145 (Aug 2nd). He may know if the problem
> has been fixed and this simply has been overlooked or if it's still there.
>
>
> --
> Jim Segrave address@hidden
>
>
> _______________________________________________
> Bug-gnubg mailing list
> address@hidden
> http://mail.gnu.org/mailman/listinfo/bug-gnubg