Re: [Bug-gnubg] Crash running command file from gui.

[Ian Shaw]

I've got this working. Thanks for all your help, and the tutorial, Michael!

I found that to get the board position correct, I had to use 4HPwATDgc+EBIg 
rather than 4HPhASLgc/ABMA. 

If you manually play Player 0 an initial 43: 24/21 13/9, then the position you 
get if you copy at the point of Player 1 being on roll is:
Position ID: 4HPhASLgc/ABMA
Match ID: cAkLAAAAAAAE

However, if you Analyse Game, then select the 43 and do Show, you see 
4HPwATDgc+EBIg, and this is what I had to use. 
 
So my Python script has:
board = '4HPwATDgc+EBIg'

        # Setup the position and get the computer's choices.
        gnubg.command ('set board %s' % (board))
        gnubg.command ('set turn 1')
        gnubg.command ('set dice %d %d' % (die1, die2))

It's not clear to me why you should need to do this, but it seems to be 
necessary, and not easy to use.

One further thing. I've noticed that if I'm rolling out a bunch of CMarked 
positions with gnubg minimised,  then the gnubg main and rollout windows 
un-minimise whenever the next rollout starts. This does not seem like desirable 
behaviour.

-- Ian

> -----Original Message-----
> From: Michael Petch [mailto:address@hidden
> Sent: 03 June 2013 19:01
> To: Ian Shaw
> Cc: address@hidden
> Subject: Re: [Bug-gnubg] Crash running command file from gui.
> 
> On 03/06/2013 10:38 AM, Ian Shaw wrote:
> >  Ah! I see what you mean now. Good spot Philippe; I hadn't noticed.
> >
> 
> Yes, the downside to setting matchid is that it resets the game list. So if 
> you are
> trying to keep all the results in a single file then your method would have 
> to be
> used. If that is what you had in mind, I amended my previous script (rather
> trivial change) to do what you were looking for as well.
> 
> --
> Michael Petch
> GNU Backgammon Developer
> OpenPGP FingerPrint=D81C 6A0D 987E 7DA5 3219 6715 466A 2ACE 5CAE 3304