[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [gnugo-devel] avoid .sgf problem
From: |
Paul Pogonyshev |
Subject: |
Re: [gnugo-devel] avoid .sgf problem |
Date: |
Sat, 21 Aug 2004 03:28:01 +0300 |
User-agent: |
KMail/1.6.52 |
Arend wrote:
> We have a .sgf-problem:
>
> Handicap stones should be set up with AB[]. If gnugo detects a game with
> handicap, but no AB[]'s, it adds them (after the first move).
It should not, agreed.
> KGS uses B[] instead of AB[] in case of free handicap placement. This may
> not be according to the letter of the sgf-specification, but I think it is
> correct according to its spirit -- AB[] is for preset stones (starting
> position
> of a problem, handicap stones), B[] is for moves that are an actual decision
> by one of the players.
>
> In fact, it seems to me that the text in the description of the HA[]
> property was only thinking about fixed handicap.
Here I don't agree. I read the specification and I clearly remember that
it said something like "HA[] doesn't mandate handicap stones placement"
i.e. HA[9] doesn't mean 9 stones at hoshi points. KGS is doing bad job
actually.
> I suggest to avoid this issue by only manually adding the handicap stones
> if we detect a _white_ move on the empty board in a handicap game.
If I were you, I'd remove this special case completely.
Paul