[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Bug-gnubg] SGG import issues
From: |
Ned Cross |
Subject: |
Re: [Bug-gnubg] SGG import issues |
Date: |
Tue, 17 Dec 2002 15:01:49 -0800 |
I've experienced a few quirks as well with .sgg files. Upon investigation,
I've found the .sgg file itself to be the culprit.
For example: In a recent game GNUbg tells me I should have doubled while on
the bar. I danced (.sgg notation is " 63: 0-0 ") and GNUbg says I danced a
second time, assigning me 2 huge blunders for missed cube action, thus
ranking my cube skill as "awful!".
Investigating further, I see that the .sgg file actually lists the "63: 0-0"
line twice, while other places in the match a dancing roll is only listed
once. Removing the duplicate line in the .sgg and re-running the analysis
fixes the problem.
Another (and very strange) example occurred when the .sgg mysteriously
"lost" most of the moves of the game, showing a Double (Very Bad!)/Pass
(Very Bad!) on the fourth roll, when the game actually went on for another
10 rolls or so before a proper racing Double/Pass. Again this was a bug in
the gamesgrid client saving the .sgg.
Looking at your attached file, I see you have a similar problem, and it's
easy to fix:
Game 7. 6-7
1 43:
1 43:
1 43: 13/9, 13/10
1 11:
SubtleOne: a sec pls
1 11: 6/5(2), 8/7(2)
your first roll of 43 is listed on 3 lines. Just cut the first line, and
re-save the .sgg. Your problem will be fixed.
You see below, that a 'normally recorded' game starts with the first roll
listed only twice. Once to indicate which player on roll (via the
indentation), and to indicate a dice roll vice a doubling cube play. The
second to indicate the checker movement.
Game 6. 4-7
1 41:
1 41: 13/9, 24/23
1 43:
1 43: 13/9, 24/21
While I wouldn't say corrupted .sgg's happen frequently, they do happen
often enough that if GNUbg is acting strange that's the first thing I check.
Ned
----- Original Message -----
From: "Albert Silver" <address@hidden>
To: "'Oystein Johansen'" <address@hidden>; "'Ian Shaw'"
<address@hidden>
Cc: <address@hidden>
Sent: Tuesday, December 17, 2002 10:46 AM
Subject: [Bug-gnubg] SGG import issues
> Hi,
>
> I was analyzing a money game session with GNU and was amazed to
> see it judge myt opponent and I as rank beginners. It's possible, but I
> didn't think we had played quite THAT bad. I then noticed that it did
> not register the first move in game 7, so that just about everything
> following it was judged a colossal blunder. I tried importing it twice
> and had the same result. I'm sending the file to you. If I were you, I'd
> ONLY look at this 7th game. No need to subject yourself to brain damage
> at such an early age. :-)
>
> Actually, the following is a worthy anecdote. Guess how my very
> first match with money on the line started? I rolled 51, and split. I
> then never made another move and was gammoned. I kid you not. The timing
> of this was so ludicrous, that even I had to smile.
>
> Albert
>
>
----------------------------------------------------------------------------
----
> _______________________________________________
> Bug-gnubg mailing list
> address@hidden
> http://mail.gnu.org/mailman/listinfo/bug-gnubg
>
- [Bug-gnubg] World Class ++ and user definitions, Albert Silver, 2002/12/17
- RE: [Bug-gnubg] World Class ++ and user definitions, Øystein Johansen, 2002/12/17
- RE: [Bug-gnubg] World Class ++ and user definitions, Øystein Johansen, 2002/12/17
- RE: [Bug-gnubg] World Class ++ and user definitions, Ian Shaw, 2002/12/17
- RE: [Bug-gnubg] World Class ++ and user definitions, Øystein Johansen, 2002/12/17
- Re: [Bug-gnubg] SGG import issues, Joern Thyssen, 2002/12/18