TalkChess.com

BrendanJNorman wrote:Do you know what would be cool? If there was a team-created GUI as their is a team-created Stockfish.

How good could a GUI become with a dedicated, worldwide team contributing? Would be awesome.

Actually this is what XBoard has been for nearly 3 decades. The problem is that the current team is just me and Josh Pettus.

hgm wrote:

BrendanJNorman wrote:Do you know what would be cool? If there was a team-created GUI as their is a team-created Stockfish.

How good could a GUI become with a dedicated, worldwide team contributing? Would be awesome.

Actually this is what XBoard has been for nearly 3 decades. The problem is that the current team is just me and Josh Pettus.

This is a really challenging problem. PyChess is a team-created GUI also (see contributors page https://github.com/pychess/pychess/graphs/contributors), but it's very hard to get new developers/contributors to a project like a chess GUI where the result is not so measurable in contrast to a chess engine where if your engine is the best, everyone wants to contribute.

gbtami wrote:

hgm wrote:

BrendanJNorman wrote:Do you know what would be cool? If there was a team-created GUI as their is a team-created Stockfish.

How good could a GUI become with a dedicated, worldwide team contributing? Would be awesome.

Actually this is what XBoard has been for nearly 3 decades. The problem is that the current team is just me and Josh Pettus.

This is a really challenging problem. PyChess is a team-created GUI also (see contributors page https://github.com/pychess/pychess/graphs/contributors), but it's very hard to get new developers/contributors to a project like a chess GUI where the result is not so measurable in contrast to a chess engine where if your engine is the best, everyone wants to contribute.

Many pieces need to fall into place for a true community effort to happen:

* The project needs to be somewhat regularly maintained; otherwise it's considered dead.
* The project needs to be well-documented and/or have a good codebase that third parties can start improving on their own.
* There has to be developers willing to donate their valuable time, they have to be interested in the very small niche of computer chess and they have to be somewhat experienced in the programming language the project is written in.
* It must be easy to contribute and discuss changes and manage issues. Put your stuff on Github!

I consider us (me and Arto) extremely lucky that we got at least a couple of skilled contributors. I never expected it and never counted on it.

Hi Ilari,
I use Cutechess-cli for testing my engine gogobello. I love it, it's very powerful for this purpose. My best congrats!

The only thing I would like to see implemented is the possibility to stop a test session (thousands of games) and resume it when I want. Do you think to add this feature in the future?

CuteChess 1.0.0 64-bit with manual adjudication.

Game > Adjudicate as White, Black, Draw

cutechess.exe
http://www86.zippyshare.com/v/7xr6gmPd/file.html

[pgn][Event "?"]
[Site "CuteChess | i7-5960X 4.1GHz"]
[Date "2017.08.25"]
[Round "?"]
[White "komodo-11.2.2-64bit"]
[Black "Houdini 5"]
[Result "0-1"]
[FEN "Nnkb3r/np1b3p/p3pp2/3p4/1P6/2P2N2/1P3PPP/R3KB1R w KQ - 2 1"]
[PlyCount "44"]
[SetUp "1"]
[Termination "adjudication"]
[TimeControl "600+10"]

1. O-O-O {-1.11/30 35s} Nbc6 {+1.08/25 33s} 2. c4 {-1.26/30 12s}
Kb8 {+1.08/26 23s} 3. b5 {-1.21/32 32s} Nb4 {+1.13/24 3.3s}
4. cxd5 {-1.14/32 17s} exd5 {+1.29/25 11s} 5. Rd4 {-1.17/33 33s}
a5 {+1.28/30 24s} 6. b6 {-1.34/31 11s} Nac6 {+1.29/29 26s} 7. Nc7 {-1.25/30 15s}
Nxd4 {+1.29/27 0.12s} 8. Nxd4 {-1.29/30 12s} Bxc7 {+1.22/30 23s}
9. bxc7+ {-1.37/29 11s} Kxc7 {+1.35/26 12s} 10. Kd2 {-1.34/29 12s}
Re8 {+1.37/24 21s} 11. Bd3 {-1.50/31 50s} Nxd3 {+1.45/25 12s}
12. Kxd3 {-1.54/29 12s} b5 {+1.40/27 23s} 13. b4 {-1.59/31 28s}
axb4 {+1.49/28 16s} 14. Rb1 {-1.53/32 22s} Kb6 {+1.54/30 26s}
15. Rxb4 {-1.57/31 21s} Rc8 {+1.61/27 18s} 16. Rb1 {-1.66/31 59s}
Rc4 {+1.61/30 49s} 17. Ra1 {-1.87/32 27s} Be8 {+1.71/29 43s}
18. f4 {-1.81/29 22s} Bd7 {+1.79/25 20s} 19. Ra8 {-1.98/30 65s}
Kc5 {+2.14/24 12s} 20. Nb3+ {-2.67/31 46s} Kb4 {+2.30/25 14s}
21. Nd4 {-2.91/32 13s} Rc3+ {+2.40/26 15s} 22. Kd2 {-3.11/32 32s}
Kc4 {+2.37/25 0.003s, Black wins by adjudication: user decision} 0-1

[/pgn]

sasachess wrote: The only thing I would like to see implemented is the possibility to stop a test session (thousands of games) and resume it when I want. Do you think to add this feature in the future?

+1

Great GUI, but this is a must. Sometimes, you need to turn off the PC or use it for other things and you can't resume the tournament later if you stop it. I hope Ilari implements this feature.

Carlos777 wrote:

sasachess wrote: The only thing I would like to see implemented is the possibility to stop a test session (thousands of games) and resume it when I want. Do you think to add this feature in the future?

+1

Great GUI, but this is a must. Sometimes, you need to turn off the PC or use it for other things and you can't resume the tournament later if you stop it. I hope Ilari implements this feature.

This is now the most requested feature and it's obvious why. I'll make it a top priority.

I just wanted to mention two problems I have with the latest version of cutechess.

1) I was excited to try out the new user adjucation but unfortunately the tournament hangs after hitting the button.(archlinux gcc-7, with qt-5.9.1)

2) Cutechess outputs a high score everytime it displays a new search depth. See the image below:

http://imgur.com/Mr2nJK8

pferd wrote:I just wanted to mention two problems I have with the latest version of cutechess.

1) I was excited to try out the new user adjucation but unfortunately the tournament hangs after hitting the button.(archlinux gcc-7, with qt-5.9.1)

2) Cutechess outputs a high score everytime it displays a new search depth. See the image below:

http://imgur.com/Mr2nJK8

Thanks for the feedback! I've logged issues about these bugs at Github: https://github.com/cutechess/cutechess/issues