Etude: first release

[Jacob]

Today's Google's 9th birthday, today's my 19th birthday, so I figured I'd make my first release of my chess engine, Etude . Its strength is probably (hopefully ) in the 2100-2300 range. I threw together a web page for it on my dad's server: rdhales.com/jacob. Once I improve the tournament-style time control and implement pondering (and maybe work on endgame eval), I'll submit it to ... Chess War? Could anyone give me more information about how that works/where I could have it play in CC tournaments?

[swami]

Could anyone give me more information about how that works/where I could have it play in CC tournaments?

Hi Jacob,

Thanks for the relase of this engine.

I can add the engine to the Edition 6,Division 8-10 Tournament at Computer Chess Blogger, it can climb its way up the ladder of higher divisions.

Regards.

[nczempin]

Today's Google's 9th birthday, today's my 19th birthday, so I figured I'd make my first release of my chess engine, Etude . Its strength is probably (hopefully ) in the 2100-2300 range. I threw together a web page for it on my dad's server: rdhales.com/jacob. Once I improve the tournament-style time control and implement pondering (and maybe work on endgame eval), I'll submit it to ... Chess War? Could anyone give me more information about how that works/where I could have it play in CC tournaments?

Don't worry, people will pick it up. Is it uci or wb?
The only tourney where you actually have to submit is OpenWar, and the current edition is already running.

[Jacob]

Is it uci or wb?

Both, but the wb is ad hoc, so I'd recommend using UCI.

Oh, and my server didn't like the .bk extension on the opening book. A new one is up in zip format.

[Jacob]

One last thing, I fixed a few errors. Some threading issues occasionally caused illegal move losses in the beginning of new games (even hamsters and lime had problems with this), and I had accidentally left some eval parameters set that grossly exaggerated certain pawn features. It should play much better now. I posted a couple of 80-game matches on my site (and a match Harald set up against his engine).

Etude .10 vs. Cyrano (?): 1/2-1/2 (2min 6sec inc)
Etude .11 vs. Hamsters 0.0.6 (2349): 22-45-13, 36%
Etude .11 vs. Lime_v62 (2227): 36-39-5, 48%
*v.11 without threading fix

Matches were 1min 60sec inc using the silver suite that Dr. Hyatt posted in the programming forums, colors alternated. Ratings are based on "Super-league." Pgn can be downloaded here (hehe, if anyone's interested)

[GS]

Is it uci or wb?

Both, but the wb is ad hoc, so I'd recommend using UCI.

Oh, and my server didn't like the .bk extension on the opening book. A new one is up in zip format.

Hi Jacob,

Is it possible that the book doesn't work in UCI mode?
I was forced to give it a general book after it did not work.

Regards,
Guenther

Code: Select all

< ENGINE id name Etude v.11
< ENGINE id author Jacob Hales
< ENGINE option name Hash type spin default 8 min 1 max 256
< ENGINE option name Clear Hash type button
< ENGINE option name OwnBook type check default false
< ENGINE option name UCI_EngineAbout type string default Etude v.11 by Jacob Hales
< ENGINE uciok
> ENGINE setoption name Hash value 128
> ENGINE setoption name OwnBook value True
> ENGINE isready
< XBOARD xboard
< XBOARD protover 2
> XBOARD feature done=0
> XBOARD feature analyze=1
> XBOARD feature colors=0
> XBOARD feature draw=1
> XBOARD feature ics=1
> XBOARD feature myname="Etude_11"
> XBOARD feature name=1
> XBOARD feature pause=0
> XBOARD feature ping=1
> XBOARD feature playother=1
> XBOARD feature reuse=1
> XBOARD feature san=0
> XBOARD feature setboard=1
> XBOARD feature sigint=0
> XBOARD feature sigterm=0
> XBOARD feature time=1
> XBOARD feature usermove=1
> XBOARD feature variants="normal"
< XBOARD accepted done
< XBOARD accepted analyze
< XBOARD accepted colors
< XBOARD accepted draw
< XBOARD accepted ics
< XBOARD accepted myname
< XBOARD accepted name
< XBOARD accepted pause
< XBOARD accepted ping
< XBOARD accepted playother
< XBOARD accepted reuse
< XBOARD accepted san
< XBOARD accepted setboard
< XBOARD accepted sigint
< XBOARD accepted sigterm
< XBOARD accepted time
< XBOARD accepted usermove
< XBOARD accepted variants
< ENGINE readyok
> XBOARD feature done=1
< XBOARD accepted done
< XBOARD new
POLYGLOT NEW GAME
POLYGLOT WAIT
> ENGINE ucinewgame
< XBOARD random
< XBOARD ics -
< XBOARD level 40 15 0
< XBOARD post
< XBOARD hard
POLYGLOT WAIT
< XBOARD easy
POLYGLOT WAIT
< XBOARD ping 1
> XBOARD pong 1
< XBOARD force
POLYGLOT WAIT
< XBOARD computer
< XBOARD name Dirty Jul 20 2007
< XBOARD time 90000
< XBOARD otim 87857
< XBOARD usermove b1c3
POLYGLOT MOVE Nc3
POLYGLOT WAIT
< XBOARD go
POLYGLOT THINK
POLYGLOT START SEARCH
POLYGLOT FEN rnbqkbnr/pppppppp/8/8/8/2N5/PPPPPPPP/R1BQKBNR b KQkq - 1 1
> ENGINE position startpos moves b1c3
> ENGINE go wtime 878570 btime 900000 movestogo 40
< ENGINE info depth 1 score cp -2 time 241 nodes 43 nps 43 pv e7e5
> XBOARD 1 -2 24 43 e5
< ENGINE info depth 2 score cp -5 time 241 nodes 158 nps 158 pv e7e5 e2e4
> XBOARD 2 -5 24 158 e5 e4
< ENGINE info depth 3 score cp -14 time 241 nodes 1345 nps 1345 pv e7e5 e2e4 d8e7
> XBOARD 3 -14 24 1345 e5 e4 Qe7
< ENGINE info depth 4 score cp -7 time 261 nodes 5371 nps 5371 pv b8c6 e2e4 e7e5
> XBOARD 4 -7 26 5371 Nc6 e4 e5
< ENGINE info depth 5 score cp -14 time 291 nodes 18788 nps 18788 pv g8f6 d2d4 d7d5 d1d2 d8d7
> XBOARD 5 -14 29 18788 Nf6 d4 d5 Qd2 Qd7
< ENGINE info depth 6 score cp -5 time 361 nodes 46123 nps 46123 pv d7d5 d2d4 g8f6
> XBOARD 6 -5 36 46123 d5 d4 Nf6
< ENGINE info depth 7 score cp -15 time 571 nodes 142076 nps 142076 pv d7d5 d2d4 g8f6 g1f3 b8c6 d1d2 d8d7
> XBOARD 7 -15 57 142076 d5 d4 Nf6 Nf3 Nc6 Qd2 Qd7
< ENGINE info currmove e7e5 currmovenumber 2 time 1002 nodes 321134 nps 320493
< ENGINE info currmove b8c6 currmovenumber 3 time 1142 nodes 379363 nps 332191
< ENGINE info currmove g8f6 currmovenumber 4 time 1172 nodes 391555 nps 334091
< ENGINE info currmove d7d6 currmovenumber 5 time 1252 nodes 424337 nps 338927
< ENGINE info currmove e7e6 currmovenumber 6 time 1282 nodes 435618 nps 339795
< ENGINE info currmove b7b6 currmovenumber 7 time 1302 nodes 444505 nps 341401
< ENGINE info currmove c7c6 currmovenumber 8 time 1312 nodes 446431 nps 340267
< ENGINE info currmove g7g6 currmovenumber 9 time 1322 nodes 451927 nps 341850
< ENGINE info currmove c7c5 currmovenumber 10 time 1332 nodes 455034 nps 341617
< ENGINE info currmove g7g5 currmovenumber 11 time 1372 nodes 473477 nps 345099
< ENGINE info currmove a7a6 currmovenumber 12 time 1382 nodes 476107 nps 344505
< ENGINE info currmove h7h6 currmovenumber 13 time 1382 nodes 477388 nps 345432
< ENGINE info currmove a7a5 currmovenumber 14 time 1382 nodes 478369 nps 346142
< ENGINE info currmove h7h5 currmovenumber 15 time 1392 nodes 479757 nps 344653
< ENGINE info currmove b8a6 currmovenumber 16 time 1392 nodes 481119 nps 345631
< ENGINE info currmove g8h6 currmovenumber 17 time 1392 nodes 481943 nps 346223
< ENGINE info currmove f7f6 currmovenumber 18 time 1392 nodes 482823 nps 346855
< ENGINE info currmove f7f5 currmovenumber 19 time 1402 nodes 483073 nps 344559
< ENGINE info currmove b7b5 currmovenumber 20 time 1402 nodes 483920 nps 345164
< ENGINE info depth 8 score cp -11 time 1402 nodes 486796 nps 347215 pv d7d5 d2d4 b8c6 d1d3 d8d7 g1f3 c6b4 d3d2
> XBOARD 8 -11 140 486796 d5 d4 Nc6 Qd3 Qd7 Nf3 Nb4 Qd2

[Jacob]

I see both UCI and xboard commands. Switching from UCI to xboard will reset the engine options,... Otherwise, as long as the book.bk file is in the same directory as the engine executable, it should work in UCI mode (and I think is used automatically in xboard).

[GS]

I see both UCI and xboard commands. Switching from UCI to xboard will reset the engine options,... Otherwise, as long as the book.bk file is in the same directory as the engine executable, it should work in UCI mode (and I think is used automatically in xboard).

The xboard options are from Polyglot as I ran it as UCI in WB.
The bookfile was of course in the same folder. BTW why is the
default UCI option with own book off?
The bookfile has around 17MB unzipped and is called book.bk.
Any idea?

Guenther(RWBC)

Edited: running it as xboard is no choice as you write at your site
that the time management is broken under it.

[Jacob]

Figured it out, parsing error in my program. It's case sensitive for the options and expects "true" instead of "True." OwnBook is false by default because I don't use books for the majority of my tests (and Arena sets it to true otherwise).

I'll have it patched probably by the end of the day.

[GS]

Figured it out, parsing error in my program. It's case sensitive for the options and expects "true" instead of "True." OwnBook is false by default because I don't use books for the majority of my tests (and Arena sets it to true otherwise).

I'll have it patched probably by the end of the day.

Thanks! That was a fast one to find ;)

Guenther