gnugo-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [gnugo-devel] sgf patch fixes


From: Gunnar Farneback
Subject: Re: [gnugo-devel] sgf patch fixes
Date: Sun, 08 Sep 2002 22:40:09 +0200
User-agent: EMH/1.14.1 SEMI/1.14.3 (Ushinoya) FLIM/1.14.2 (Yagi-Nishiguchi) APEL/10.3 Emacs/20.7 (sparc-sun-solaris2.7) (with unibyte mode)

Paul wrote:
> about --printsgf:
>   - do we need filename after this option? doesn't -o option do
>     here?

No strong opinions on my part but it's slightly clumsy to have to
write "--printsgf -o game.sgf" instead of "--printsgf game.sgf".

>   - should i add output of PL and IL tags as it was in previous
>     implementation?

PL tag yes. I don't know if PL previously could be affected by a
--color option, but that would definitely be desirable.

I don't like the IL property but presumably the cgi stuff in
interface/html/* would break without it. On the other hand, did it
still work in 3.2 and do we want to continue supporting it in either
case?

> If one uses --printsgf to generate a diagram to be used in a regression
> test, then these look necessary to me -- I guess that's why they were
> invented in the first place.

The IL property can in this context be used to simulate that a move is
forbidden due to ko. However, in such cases it is much better to do
printsgf at the previous move and add the last one as an explicit
move. Since IL is a private property we can't expect other programs to
understand it, but we do want our regressions to be generally useful.
Thus we should avoid it.

/Gunnar




reply via email to

[Prev in Thread] Current Thread [Next in Thread]