New Unix client & SGF editor
William M. Shubert
connect to a go server as a client program, it can view and edit smart-go files, it can act
as a "bridge" to the go modem protocol letting you connect computer programs to go servers,
and you can just play go with it.
It includes (IMHO) good graphics with rendered stones, a texture mapped board, etc.
Please look at the wex page:
http://www.hevanet.com/wms/comp/cgoban/
Trust me! It's really cool!
As an added bonus to Bill Taylor, when you are playing locally it applies STRICT Japanese
rules. This means that it recognizes Seki positions and awards no points, and it has a
system for resolving whether disputed stones are alive or dead at the end of the game by
playing it out on a separate board with slightly different ko rules.
--
-Bill (w...@hevanet.com)
http://www.hevanet.com/wms/
Dan Putnam
: I have written a Unix/X11 program that does a lot of useful go-related things. It can
: connect to a go server as a client program, it can view and edit smart-go files, it can act
: as a "bridge" to the go modem protocol letting you connect computer programs to go servers,
: and you can just play go with it.
: It includes (IMHO) good graphics with rendered stones, a texture mapped board, etc.
: Please look at the wex page:
: http://www.hevanet.com/wms/comp/cgoban/
: Trust me! It's really cool!
Yep, after reading some of the other responses and visiting your
web page, I believe you! Unfortunately, I don't have a Unix
system that I can use to try it out :(
Now, if we could just get you interested in Java...
Erik Van Riper
William M. Shubert <w...@hevanet.com> wrote:
>I have written a Unix/X11 program that does a lot of useful go-related things. It can
>connect to a go server as a client program, it can view and edit smart-go files, it can act
>as a "bridge" to the go modem protocol letting you connect computer programs to go servers,
>and you can just play go with it.
> It includes (IMHO) good graphics with rendered stones, a texture mapped board, etc.
>Please look at the wex page:
>
> http://www.hevanet.com/wms/comp/cgoban/
>
> Trust me! It's really cool!
Having seen this, and playing with the sgf editor, I have to admit, it is
absolutly fantastic. :)
Great job, Bill!
--
ge...@imageek.york.cuny.edu http://imageek.york.cuny.edu
Erik Van Riper (EV34) (718) 262-2667
Systems Administrator Go player Photon Counter Coffee lover
Language design is 10% science and 90% psychology. -- Larry Wall
Henrik Hamelius
In article <4cb6jl$q...@news.cuny.edu>, ge...@imageek.york.cuny.edu (Erik Van Riper) writes:
|> In article <30E8CDC6...@hevanet.com>,
|> William M. Shubert <w...@hevanet.com> wrote:
|> >I have written a Unix/X11 program that does a lot of useful go-related things. It can
|> >connect to a go server as a client program, it can view and edit smart-go files, it can act
|> >as a "bridge" to the go modem protocol letting you connect computer programs to go servers,
|> >and you can just play go with it.
|> > It includes (IMHO) good graphics with rendered stones, a texture mapped board, etc.
|> >Please look at the wex page:
|> >
|> > http://www.hevanet.com/wms/comp/cgoban/
|> >
|> > Trust me! It's really cool!
|>
|> Having seen this, and playing with the sgf editor, I have to admit, it is
|> absolutly fantastic. :)
|>
|> Great job, Bill!
|>
|> --
|> ge...@imageek.york.cuny.edu http://imageek.york.cuny.edu
Just has to agree! Wow!!
/hamlet
Arnold Jonk
>|> Having seen this, and playing with the sgf editor, I have to admit, it is
>|> absolutly fantastic. :)
>|>
>|> Great job, Bill!
>|>
>|> --
>|> ge...@imageek.york.cuny.edu http://imageek.york.cuny.edu
>Just has to agree! Wow!!
>/hamlet
:( :(
I can't get it to compile. I get this silly `don't know how to make
libclient' error, and don't think I did anything wrong. Can anybody
help me out with a binary or advice?
Arnold
--
I put my feet where the earth should be.
-- Heather Nova --
Nick Wedd
>I have written a Unix/X11 program that does a lot of useful go-related things.
< snip.. >
>it recognizes Seki positions and awards no points
I think that no other program has achieved this. Some of them recognise
some sekis, but get confused by the status of others. Congratulations!
Nick.
--
Nick Wedd Ni...@maproom.demon.co.uk 72133...@compuserve.com
Bill Newman
: "William M. Shubert" <w...@hevanet.com> writes
: >I have written a Unix/X11 program that does a lot of useful go-related things.
: < snip.. >
: >it recognizes Seki positions and awards no points
: I think that no other program has achieved this. Some of them recognise
: some sekis, but get confused by the status of others. Congratulations!
I don't know about other programs, but I know that I tried to do this
in my program and failed. I thought I was doing OK when I figured out
how to recognize things like the mess in the lower right of the
drawing below, but then I realized that groups can have an arbitrary
number of liberties and still be seki, like the mess in the center, at
which point I gave up. I'd be fascinated to see a really general
algorithm, or, short of that, a guaranteed-complete bestiary of
classes of seki shapes.
. . . . . . . . . . . . . . . . . .
. . . . . # # # O O . . . . . . . .
. . . . . # O O # O . . . . . . . .
. . . . . # O . # O O O O . . . . .
. . . # # # O . # # # # O . . . . .
. . . # O O O O # . . # O . . . . .
. . . O # . . # O O O O # . . . . .
. . . O # # # # . O # # # . . . . .
. . . O O O O # . O # . . . . . . .
. . . . . . O # O O # . . . . O O O
. . . . . . O O # # # . . . . O # #
. . . . . . . . . . . . . . . O # .
. . . . . . . . . . . . . . O O # #
. . . . . . . . . . . . . . O # O .
. . . . . . . . . # # # # O O # O O
. . . . . . . . . # O O O # # # # O
. . . . . . . . . # O . O . # . O O
Bill Newman
wne...@netcom.com
Charlie Wine
The high quality shell and slate stones are most pleasing.
Thanks for investing the effort (must be considerable) to
create a fine new go client.
William M. Shubert
> Nick Wedd (Ni...@maproom.demon.co.uk) wrote:
> : "William M. Shubert" <w...@hevanet.com> writes
>
> : >I have written a Unix/X11 program that does a lot of useful go-related things.
>
> : < snip.. >
>
> : >it recognizes Seki positions and awards no points
>
> : I think that no other program has achieved this. Some of them recognise
> : some sekis, but get confused by the status of others. Congratulations!
>
> I don't know about other programs, but I know that I tried to do this
> in my program and failed.
> I'd be fascinated to see a really general
> algorithm, or, short of that, a guaranteed-complete bestiary of
> classes of seki shapes.
Since I was only identifying seki for the purposes of scoring, it became much easier.
The hard part of identifying seki is spotting dead groups. Cgoban already knows exactly
which stones are alive and dead, because the users have either told the program or else
resolved through the dispute mechanism of cgoban which groups are alive. From here, cgoban
just fills in the dame (filling fake eyes as needed), then looks at what is left. If a
shape has only one eye, and can not make a second, then it is seki. Figuring out whether
you can make a second eye is pretty simple, you just need to recognize a few shapes.
The only problem is that it sometimes cannot fill the dame correctly. For example:
# # #
# O O O
# . # O
# . # O O O O
O O O # # # O O #
. O . # . # . . #
O O # # # # O O #
Here cgoban will get confused. Cgoban fills dame with "grey" stones that consume
liberties but also connect through for both white and black. Can the two one-eyed shapes
connect with their outside groups? Cgoban will think that they can, and incorrectly decide
that they are not in seki. If you are playing cgoban and you see this problem (it will be
clear because the two seki eyes will be marked as points), then you must continue the game
and fill in the dame in the bamboo joints. Then go back to scoring and they will be
correctly marked as seki.
This doesn't bother me too much since in tournament games with japanese rules, the dame
_are_ filled in at the end of the game. If there is a scoring error in cgoban, just fill in
the dame and it should become correct. Errors like this are extremely rare anyway (sekis
are rare, sekis with eyes are rarer, and a position like the one above I have never seen in
an actual game).
Needless to say, if the players make mistakes and mark living stones as dead or dead
stones as living then cgoban can get _very_ confused when it tries to find sekis.