[Bug-gnubg] None

[Hanley Hayes]

Running the Windows Dec 27 build, with the 2-sided bearoff DB. I have the 
following four bugs/questions.  The first two are from the attached 13 point 
match.

1. Game 1.  Move 23.  I have one checker on my one point and one checker on 
my two point.  GNU has one checker on his six point and is on roll.  He 
rolls 52 and wins the game.  Move is marked as "lucky".  As I see it, the 
only rolls that wouldn't win immediately for GNU are 11, 12, 21, 13, 31, 14, 
41, 23, 32 or 9/36.  Why is a 27/36 roll marked as "lucky"?

2. Game 11.  Move 20.  GNU resigns when he is mathematically unable to win.  
Yet the annotation window says: Resign - Very Bad,  Accept - Very Bad.

3. Is the Windows executable 'debuggable', or is what I downloaded from 
//users/skynet.be a version compiled without debug on?  The reason I ask is 
that about once a week gnubg crashes on me (no particular pattern, for 
example, most recently while analyzing a match) and I would be happy to send 
you more info, but there is no core file on my disk.  While the product is 
still slightly unstable, would it be possible to put on the website a 
version that has debug on, so I would something to send you?  I realize that 
I could build my own ... but would rather not.

4. At least three times now, I have gotten the Game Records window's Game 
Score Combo Box/Option Menu/Pulldown Menu to be out of sync with what the 
Game Score window is actually showing me - for example, it will say, Game 
10: GNU 11 Hanley 8, when the game shown is actually game 12.  I don't know 
what I did to get in this 'mode', and it was easy to correct (select a 
different game from the menu and then return to the desired one), but I'm 
wondering if this has been reported before.



_________________________________________________________________
MSN 8: advanced junk mail protection and 2 months FREE*. 
http://join.msn.com/?page=features/junkmail

2003-01-03-GNU-Hanley-13.sgf
Description: Binary data