Gipf Project

[ujh]

The Gipf Project is a set of 6 abstract board games (Gipf, Tamsk/Tzaar, Zertz, Dvonn, Yinsh, Pünct) that can be played separately or tied together by the game called Gipf.

The main page of the games is http://www.gipf.com/

Various Links:

• http://en.wikipedia.org/wiki/GIPF_project
• Dvonn can be played online at http://www.littlegolem.net/jsp/games/ga ... gtid=dvonn . There is at least one program that plays there: http://www.littlegolem.net/jsp/info/pla ... ?plid=3107
• All Gipf Project games can be played by email on Richard's PBeM Server: http://www.gamerz.net/pbmserv/ . Although no ideal format it should be possible to automatically connect a program as the emails can be parsed.

[ddyer]

Zertz Yinsh and Punct can be played at Boardspace.net. There are modestly good robot
players for all three; good enough to trounce beginners, but intermediate level players
will have no problem beating the robots.

[vintermann]

One more useful link:

Kurt van den Branden's program GF1 is the computer world champion (and human, too). It's open source, and can be found here:

http://gf1.sourceforge.net

Unfortunately, it hasn't been updated for a long time, and time is rarely kind to old c++-programs... I can't get it to compile, and I have no clue how to update it to c++ that today's gcc will accept.

I did get the binary to work, but only through some dubious symlinking.

What would be extremely sweet was if someone wanted to do for gf1 what ab did for six: Split the UI from the AI, and give both a GTP interface. Unfortunately it seems to be a lot of work!

[MartinMPedersen]

Kurt van den Branden's program GF1 is the computer world champion (and human, too). It's open source, and can be found here:

http://gf1.sourceforge.net

Unfortunately, it hasn't been updated for a long time, and time is rarely kind to old c++-programs... I can't get it to compile, and I have no clue how to update it to c++ that today's gcc will accept.

Yes the code needs some updating and shouldn't be that hard.

adding 'using name std' to the header-files and changing <hash_map> to <map> removed many of the errors but it
does need updating a bit more. It shouldn't take more than a few hours.

The windows binary ran without problems under wine.

/Martin

[ujh]

Yes the code needs some updating and shouldn't be that hard.

adding 'using name std' to the header-files and changing <hash_map> to <map> removed many of the errors but it
does need updating a bit more. It shouldn't take more than a few hours.

I started doing that and tried to fix the easy problems. I've set up a darcs repository.

Code: Select all

darcs get http://darcs.bettong.net/gf1

Any help is of course welcome.

Urban

[Rémi Coulom]

Hi,

I got it to compile and run. I am not familiar with darcs, so I am not sure how I should post a patch. In attachment is a file generated by "darcs record", but that may not be what you want.

I am running Ubuntu 6.06. I had to install libfltk-devel and libgd-devel.

.d files are dependencies generated by the makefile. It may be better not to include them in the repository. I regenerated them by deleting them and typing "make".

Rémi

[ujh]

Hi,

I got it to compile and run. I am not familiar with darcs, so I am not sure how I should post a patch. In attachment is a file generated by "darcs record", but that may not be what you want.

I am running Ubuntu 6.06. I had to install libfltk-devel and libgd-devel.

.d files are dependencies generated by the makefile. It may be better not to include them in the repository. I regenerated them by deleting them and typing "make".

Rémi

Thanks Rémi!

Something was missing from the patch file (I think you are supposed to use "darcs send -o patch.txt" after doing a "darcs record") so I couldn't apply it via darcs. But the changes were few enough so that I just did it by hand.

So GF1 now compiles on my machine, too (Mac OS X 10.5.2, FLTK and GD installed via Fink).

Urban

[vintermann]

ujh, thanks a lot for rescuing an old champion app! I got it compile, after fixing a small error at line 40 of drawgif.h (my gcc didn't like the explicit qualification of pos2coor)