Re: [Bug-gnubg] gnubg python problems

[Martin Janke]

On Sat, 2003-11-29 at 14:29, Joseph Heled wrote:
> Martin Janke wrote:
... 
> > 1. all information in gnubg.match() is swapped, i.e. eveything belonging
> > to player 'X' can be found under 'O' and vice versa.
> > 
> Not sure what you mean. Each player has a name, and the mapping X <-> 
> player name 1, O <-> player name 2 is given in the match info.
> (i.e. gnubg.match()['match-info']['X']['name'])

You're right. gnubg.match() is correct and consistent in itself. The
only minor issue is that 'X' and 'O' are swapped in regards to the
boards in tty mode, as in this example:

   GNU Backgammon  Position ID: NAAAEAAAAAAAAA
                    Match ID   : MIEmAAAAAAAA
    +-1--2--3--4--5--6-------7--8--9-10-11-12-+  O: gnubg
OOO | O                |   |                  |  0 points
OOO |                  |   |                  |  Rolled 51
OOO |                  |   |                  |
OOO |                  |   |                  |
 OO |                  |   |                  |
    |                  |BAR|                  |^ 1 point match (Cube: 1)
 XX |                  |   |                  |
 XX |                  |   |                  |
 XX |                  |   |                  |
XXX |          X       |   |                  |
XXX |       X  X       |   |                  |  0 points
    +24-23-22-21-20-19------18-17-16-15-14-13-+  X: You

gnubg moves 1/off.
gnubg wins a single game and 1 point.
The score (after 1 game) is: gnubg 1, You 0 (match to 1 point,
post-Crawford play).
gnubg has won the match.
(Game over) >
Python 2.2.3 (#1, Nov  4 2003, 20:44:52)
[GCC 3.2.3 20030422 (Gentoo Linux 1.4 3.2.3-r2, propolice)]
>>> gnubg.match()['match-info']
{'match-length': 1, 'O': {'name': 'You'}, 'date': (29, 11, 2003), 'X':
{'name': 'gnubg'}, 'default-eval-context': {'plies': 0, 'deterministic':
1, 'reduced': 0, 'noise': 0.0, 'cubeful': 1}, 'variation': 'Standard'}
>>>

/Martin

-- 
Martin Janke <address@hidden>