SlowChess Blitz Classic 2.0

[jonkr]

CMCanavessi :

There is no console mode built into the windows version. To create the standard console is only automatic when built for it, so never bothered. The Linux version is compiled for console-mode only though.

So the better question seems to be (and I also had this question when I tested with SlowChess): How does a GUI like Arena manage to start the slow.exe without displaying its GUI but instead get access to its stdin and stdout streams. And how can a Windows user do this with Windows tools? Maybe not possible.

Regards, Andreas

On startup Slow checks if there is a pipe setup with fgets on stdin, if there is, it doesn't call CreateWindow and setup the own GUI window. So if an external UI (eg. Arena) supports sending commands to engines you could interact via uci commands that way. (Without a GUI communicating there would be no way to interact with the engine if it didn't show the own GUI).

[CMCanavessi]

CMCanavessi :

There is no console mode built into the windows version. To create the standard console is only automatic when built for it, so never bothered. The Linux version is compiled for console-mode only though.

So the better question seems to be (and I also had this question when I tested with SlowChess): How does a GUI like Arena manage to start the slow.exe without displaying its GUI but instead get access to its stdin and stdout streams. And how can a Windows user do this with Windows tools? Maybe not possible.

Regards, Andreas

On startup Slow checks if there is a pipe setup with fgets on stdin, if there is, it doesn't call CreateWindow and setup the own GUI window. So if an external UI (eg. Arena) supports sending commands to engines you could interact via uci commands that way. (Without a GUI communicating there would be no way to interact with the engine if it didn't show the own GUI).

Thanx for the explanation. So here's request time x 2

a) can you add an option to start it in console or uci mode? -c or -uci or whatever
b) when the engine loads, i see the following

info string 9 bitbases loaded.
info string 14 neural nets loaded.
info string 34237 opening positions loaded.
readyok

it will show "9 bitbases loaded" wether I use syzygy or not, which is fine because the bitbases file is in the engine's directory. But it would be nicer (maybe) if when using syzygy, it would print another message like "xxx syzygy bases loaded" or something like that, just to be sure they are being correctly found and loaded

And another question, if both bitbases and syzygy are loaded, does it use both? if not, which ones are used? and if yes, when does the engine use one or the other?

[jonkr]

CMCanavessi :

If both bitbases and tablebases are loaded Slow will use both. The Syzygy tables are queried during the main search, and the bitbases are queried on any evaluation (but they are small and only certain 4-piece or less endgames.) Not printing the syzygy loaded info string to UCI is an oversight will be fixed next version. Since having a default console window is compile-time setting and I haven't found myself wanting to use a uci command line, I'm not sure if I'll get to it, but finding a way to handle it on startup it's on my list of things to look into.

[CMCanavessi]

Thanx a lot for your reply and your patience!

[Wolfgang]

Slow 2.8NN @ CEGT-40/4
https://cegt.forumieren.com/t1621-testi ... litz-2-8nn

Games: 2000
AvOppELO: 3400 (CEGT)

[jonkr]

Slow Chess 2.9 is now available on the web page :
Slow Chess Web Page

It's not a lot stronger, In my tests when compared to 2.8 it's +15 Elo standard and +30 Elo FRC.
I just mainly wanted to release another major version instead of leaving it at 2.83, and unlike 2.83, 2.9 has always scored clearly stronger when I've played tests with a lot of games. It includes new nets and some search tweaks and speed optimizations.

I'm not planning on further updates to Slow Chess, but it's always possible at some point in the future. The rate of improvement while still there has clearly slowed down. Recently I have rarely been running tests. I like running experiments, but lately the results have always been inconclusive even if I leave my machine running chess games for 2/3rds of the day. Initially experimenting with creating a neural net trainer and evaluation function was fun and it was fun to learn more about neural nets. However the last result that actually felt exciting to me was adding the pawn-king-material hash with initial dropout training in 2.6, which was quite a long time ago. Since then mainly I kept going because further progress would become clear even when not spending a lot of time. I hope in the future there will be some exciting development in computer chess that I will feel compelled to try out. (I'm not sure what it would be, but I think it would be possible to use machine learning to apply more directly to selective search in a way that makes it easier to create a well-tuned search with less trial and error testing, but I don't have enough time for chess right now.)

[Frank Quisinsky]

Hm,

Jonathan ...
Yesterday, Wasp 5.54 NN dev won in round 3 of 30 "run-6" vs. Slow 2.83.
I believe the first won game after a very long time vs. your engine.

Not easy ...
Slow is with Wasp, Revenge and Booot ... Velvet is very nice ... my favorite engine.
John like your engine also a lot.

My new run-6 I started for 3 days.
A lot of games are over!

To add the new Slow is for me quite clear!
But I lost the nice game Wasp won yesterday ...

Hard, very hard ... one of the hardest decision in computerchess in the last years!
On the other hand ... Slow 2.9 is available ... highlight for me!

OK, if Wasp 5.54 dev NN are able to beat Slow 2.83 ... maybe it's possible in one of the 30 games vs. Slow 2.9 also?

In the next hours ...
Go Slow 2.9 go ...

Loud thinking only!

Thank you!
Frank

PS:
Waiting a time Jonathan.
The day will come and you have a great new idea to make Slow again stronger.
Never you should gave up the development from such a fantastic work you do here.
From my view ... you have the "GM title" in programmings.

[Thomas Lagershausen]

[fen]r1br2k1/1p1nqppp/p1p1pn2/3p4/1bPP4/PPN1PN2/2BB1PPP/R2Q1RK1 b - - 0 11[/fen]

SlowChess 2.9:
19/28 00:01 12.342.266 7.390.578 +0,04 Lb4xa3 e3-e4 d5xe4 Sc3xe4 Sf6xe4 Lc2xe4 La3-b4 Ld2xb4 De7xb4 Tf1-e1 Db4-e7 b3-b4 Sd7-f6 Le4-d3 b7-b6 c4-c5 Lc8-b7 Dd1-d2 h7-h6 c5xb6 Sf6-d5 Sf3-e5 De7xb4
20/30- 00:04 30.926.969 7.276.933 -0,20 Lb4xa3 Ta1xa3
20/33 00:05 38.296.519 7.212.150 0,00 Lb4xa3 Ta1xa3 De7xa3 c4-c5 Sf6-e4 Lc2xe4 d5xe4 Sf3-g5 Sd7-f6 Dd1-c2 Sf6-d5 Sc3-b1 Da3-a1 Sb1-c3 Da1-a3
21/33- 00:05 38.843.396 7.219.962 -0,24 Lb4xa3 Ta1xa3
21/33 00:06 45.189.308 7.230.289 -0,36 Lb4xa3 Ta1xa3 De7xa3 c4-c5 Sf6-e4 Lc2xe4 d5xe4 Sf3-g5 Sd7-f6 Dd1-c2 Sf6-d5 Sc3xd5 Td8xd5 Sg5xe4 e6-e5 Se4-c3 e5xd4 Sc3xd5 c6xd5 e3xd4 a6-a5 Dc2-b1 Lc8-e6 Tf1-e1 h7-h6
22/33+ 00:07 50.596.897 7.207.535 -0,12 Lb4xa3
22/33 00:07 53.630.628 7.189.092 -0,16 Lb4xa3 Ta1xa3 De7xa3 c4-c5 b7-b5 b3-b4 Da3-b2 Sf3-e1 Sf6-e4 Sc3xe4 d5xe4 f2-f3 a6-a5 f3xe4 a5-a4 e4-e5 Sd7-f8 h2-h4 Ta8-a7 Se1-d3 Db2-a2 Sd3-f2 Da2-d5 h4-h5 h7-h6 Sf2-e4 a4-a3 Se4-d6
22/35 00:10 73.719.067 7.081.562 0,00 Lb4xa3 Ta1xa3 De7xa3 c4-c5 b7-b5 b3-b4 Da3-b2 Sf3-e1 Sf6-e4 Sc3xe4 d5xe4 f2-f3 a6-a5 f3xe4 a5-a4 e4-e5 Td8-f8 Se1-d3 Db2-a2 Sd3-f4 f7-f5 Ld2-e1 Tf8-f7 h2-h4 Sd7-f8 h4-h5 Da2-b2 Sf4-d3 Db2-a2 Sd3-f4
23/35 00:11 78.348.074 7.064.749 -0,16 Lb4xa3 Ta1xa3 De7xa3 c4-c5 b7-b5 b3-b4 Da3-b2 Sf3-e1 Sf6-e4 Sc3xe4 d5xe4 f2-f3 a6-a5 f3xe4 a5-a4 e4-e5 Td8-f8 Se1-d3 Db2-a2 Sd3-f4 f7-f5 Ld2-e1 Tf8-f7 h2-h4 Sd7-f8 h4-h5 Da2-b2 e3-e4 a4-a3 e4xf5 a3-a2
24/37 00:14 103.460.288 7.019.015 +0,01 Lb4xa3 Ta1xa3 De7xa3 c4-c5 b7-b5 b3-b4 Da3-b2 Sf3-e1 Sf6-e4 Sc3xe4 d5xe4 f2-f3 a6-a5 f3xe4 a5-a4 e4-e5 Td8-f8 Se1-d3 Db2-a2 Sd3-c1 Da2-b2 Sc1-d3
25/38- 00:16 115.478.434 7.015.700 -0,23 Lb4xa3 Ta1xa3
25/43 00:20 144.823.415 7.009.845 -0,24 Lb4xa3 Ta1xa3 De7xa3 c4-c5 b7-b5 Sc3-b1 Da3-a2 Dd1-c1 Sf6-e4 Sb1-c3 Se4xc3 Ld2xc3 b5-b4 Lc3xb4 a6-a5 Lb4-c3 a5-a4 b3xa4 Lc8-a6 Dc1-d2 Da2-c4 Tf1-e1 e6-e5 Sf3xe5 Sd7xe5 d4xe5 Dc4xc5 Lc3-b4 Dc5-c4 Lb4-d6 g7-g6 f2-f3 c6-c5 Te1-c1 Dc4-e2 Dd2xe2 La6xe2 Ld6xc5
26/43 00:24 175.345.203 7.016.614 -0,36 Lb4xa3 Ta1xa3 De7xa3 c4-c5 b7-b5 Sc3-b1 Da3-a2 Dd1-c1 Sf6-e4 Sb1-c3 Se4xc3 Ld2xc3 b5-b4 Lc3xb4 a6-a5 Lb4-c3 a5-a4 b3xa4 Da2-c4 Sf3-d2 Dc4-a6 f2-f4 Da6-a7 h2-h3 Lc8-a6 Tf1-f2 Sd7-f6 Dc1-e1 Td8-e8 Tf2-f3 Kg8-f8 De1-h4 La6-e2 Tf3-f2 Da7-a6 Lc2xh7
27/43- 00:35 251.256.336 7.057.762 -0,60 Lb4xa3 Ta1xa3
27/43+ 00:39 280.758.522 7.043.615 -0,59 Lb4xc3
27/43 00:57 398.418.370 6.988.569 -0,44 Lb4xc3 Ld2xc3 b7-b6 Dd1-b1 Lc8-b7 c4xd5 e6xd5 Tf1-c1 Sf6-e4 Lc3-e1 a6-a5 b3-b4 a5xb4 a3xb4 Ta8xa1 Db1xa1 Td8-a8 Da1-b2 Se4-d6 Lc2-d3 h7-h6 Tc1-a1 De7-e8 Db2-c3 Ta8xa1 Dc3xa1 Sd6-c4 h2-h3 De8-b8 Da1-c3 b6-b5

It is so impressive to see the development of Slow from v2.6 to latest 2.9.

Hope you will find time to work further on Slow.

Till then all the best.

[Damir]

CMCanavessi :

If both bitbases and tablebases are loaded Slow will use both. The Syzygy tables are queried during the main search, and the bitbases are queried on any evaluation (but they are small and only certain 4-piece or less endgames.) Not printing the syzygy loaded info string to UCI is an oversight will be fixed next version. Since having a default console window is compile-time setting and I haven't found myself wanting to use a uci command line, I'm not sure if I'll get to it, but finding a way to handle it on startup it's on my list of things to look into.

Hi Jonathan

Is your engine UCI or winboard engine?

I have tried and create it as new engine in Fritz GUI. The author field is empty and engine does not want to be created?

I am using avx2 compile. My hardware is AMD 3990x 64 cores.

[Modern Times]

It is UCI. I haven't tried Fritz, will do so later, but it installs in CuteChess GUI just fine here.

EDIT: Also does not install here under Fritz 17 GUI.