Schooner Version 2.1

Discussion of anything and everything relating to chess playing software and machines.

Moderators: hgm, Rebel, chrisw

D Sceviour
Posts: 570
Joined: Mon Jul 20, 2015 5:06 pm

Schooner Version 2.1

Post by D Sceviour »

- WinBoard compatible, CECP compliant, ICS capable
- limited UCI protocol also available
- 64 bit version
- SSE popcnt support
- non-popcount generic available
- default hash size 128 Mb - adjustable
- default pawn hash 30 Mb - adjustable
- threading allowed - 128 threads
- pondering allowed, xboard only
- optional polyglot book - default "book.bin"
- adjustable sc.ini file
- syzygy root DTZ endgame table bases supported

https://sites.google.com/site/schoonerchess/

Schooner is an XBOARD protocol engine. It is best to use Schooner in xboard mode. There is an "sc.ini" file included so all the parameters can be sent to the command line interface. Schooner has limited UCI capability and not all features will work in UCI mode.

A BMI2 build is not included. There was no demonstrable improvement in speed for the pext instruction.

The sse version will play about 50 elo stronger than the generic non-popcount version.

Schooner supports polyglot books. Book.bin has been relegated to the old files. Arena's Perfect2017.bin opening book is now included as a starter book (108K).

(1) Prefetch instruction added for SSE build.

(2) Adjustable pawn hash size. Set phash to about 10-20% of hash size. The default pawn hash value is 30 Mb.

(3) Checking extensions for analysis mode only. This gives better mate announcements in analysis mode!

(4) Limited UCI epd and infinite support has been added.

(5) Syzygy root DTZ end game tablebase support is included. WDL probes are not used because they did not demonstrate any elo gain, especially with a six-piece egtb. The bottleneck on the main bus transfer may be intolerable. Schooner already has the basic endgames pre-programmed. They are much faster than a WDL probe.

(6) The Poor Man's KP Bitbase is included in the endgame horizon evaluation!

(7) Hopefully there will be about a 50 elo improvement over the previous release.
Gabor Szots
Posts: 1362
Joined: Sat Jul 21, 2018 7:43 am
Location: Szentendre, Hungary
Full name: Gabor Szots

Re: Schooner Version 2.1

Post by Gabor Szots »

Thank you, Dennis. 50 Elo is huge when you are already at this height.
Gabor Szots
CCRL testing group
Max
Posts: 247
Joined: Tue Apr 13, 2010 10:41 am

Re: Schooner Version 2.1

Post by Max »

Unfortunately Schooner 2.1 crashed running with wine 4 on macOS.

Code: Select all

wine64 ./Schooner2.1-sse.exe 
preloader: Warning: failed to reserve range 000000007ff00000-000000007fff0000
phash=30Mb
Schooner 2.1-sse
threads=1
hash=128MB
wine: Unhandled page fault on read access to 0x00000050 at address 0x7bc2b0ba (thread 0009), starting debugger...
Unhandled exception: page fault on read access to 0x00000050 in 64-bit code (0x000000007bc2b0ba).
Register dump:
 rip:000000007bc2b0ba rsp:000000000032f920 rbp:000000000032f950 eflags:00010202 (  R- --  I   - - - )
 rax:ffffffff95ee9020 rbx:0000000000468448 rcx:0000000000000030 rdx:000000006a116fe0
 rsi:0000000000000030 rdi:000000000032e973  r8:0000000000000000  r9:0000000000000000 r10:00007fff77d80142
 r11:0000000000000206 r12:0000000002812060 r13:000000000032e970 r14:000000000046ada0 r15:0000000000000001
Stack dump:
0x000000000032f920:  0000000500000004 0000000700000006
0x000000000032f930:  0000000100000000 0000000300000002
0x000000000032f940:  0000000100000001 0000000000000000
0x000000000032f950:  000000000032f990 000000006a0a67ed
0x000000000032f960:  0000000000000000 0000000000000000
0x000000000032f970:  0000000200000003 0000000200000002
0x000000000032f980:  000000000032e973 0000000000468480
0x000000000032f990:  0000000002812030 00000000004090ac
0x000000000032f9a0:  3ff0000000000000 0000000000000000
0x000000000032f9b0:  3fe0000000000000 0000000000000000
0x000000000032f9c0:  4058c00000000000 0000000000000000
0x000000000032f9d0:  0000000500000004 0000000700000006
Backtrace:
=>0 0x000000007bc2b0ba (0x000000000032f950)
0x000000007bc2b0ba: cmpq	$0,0x0000000000000020(%rsi)
Modules:
Module	Address					Debug info	Name (10 modules)
PE	          400000-          6ef000	Deferred        schooner2.1-sse
PE	        681d0000-        681d4000	Deferred        version
PE	        6a080000-        6a084000	Deferred        msvcrt
PE	        6a160000-        6a23a000	Deferred        user32
PE	        6a3c0000-        6a3c4000	Deferred        gdi32
PE	        6a530000-        6a534000	Deferred        advapi32
PE	        6a6a0000-        6a6a4000	Deferred        imm32
PE	        7b410000-        7b46a000	Deferred        kernel32
PE	        7bc10000-        7bc14000	Deferred        ntdll
ELF	        7c400000-        7c404000	Deferred        <wine-loader>
Threads:
process  tid      prio (all id:s are in hex)
00000008 (D) Z:\Users\max\Downloads\Schooner2.1\Schooner2.1-sse.exe
	0000002c    0
	00000009    0 <==
0000000e services.exe
	00000020    0
	0000001b    0
	00000015    0
	00000014    0
	00000013    0
	00000010    0
	0000000f    0
00000011 winedevice.exe
	00000018    0
	00000017    0
	00000016    0
	00000012    0
00000019 plugplay.exe
	0000001d    0
	0000001c    0
	0000001a    0
0000001e winedevice.exe
	00000028    0
	00000027    0
	00000026    0
	00000025    0
	00000022    0
	00000021    0
	0000001f    0
00000023 explorer.exe
	0000002b    0
	0000002a    0
	00000029    0
	00000024    0
System information:
    Wine build: wine-4.0.1
    Platform: x86_64
    Version: Windows 7
    Host system: Darwin
    Host version: 17.7.0
Is there any hope for a (native) mac version?
Hope we're not just the biological boot loader for digital super intelligence. Unfortunately, that is increasingly probable - Elon Musk
D Sceviour
Posts: 570
Joined: Mon Jul 20, 2015 5:06 pm

Re: Schooner Version 2.1

Post by D Sceviour »

Max wrote: Wed Jul 03, 2019 12:00 pm Unfortunately Schooner 2.1 crashed running with wine 4 on macOS.
Is there any hope for a (native) mac version?
Sorry, Schooner only runs on windows. I would like to create a Linux build. However, it seems like there is always one more thing to do. Day after week after month passes and an update never gets released. Are there any Windows->Mac converters besides Wine you might try?
D Sceviour
Posts: 570
Joined: Mon Jul 20, 2015 5:06 pm

Re: Schooner Version 2.1

Post by D Sceviour »

Gabor Szots wrote: Wed Jul 03, 2019 10:14 am Thank you, Dennis. 50 Elo is huge when you are already at this height.
That will be fun for you to test and find out for sure. :D
User avatar
Guenther
Posts: 4605
Joined: Wed Oct 01, 2008 6:33 am
Location: Regensburg, Germany
Full name: Guenther Simon

Re: Schooner Version 2.1

Post by Guenther »

D Sceviour wrote: Wed Jul 03, 2019 2:16 pm
Gabor Szots wrote: Wed Jul 03, 2019 10:14 am Thank you, Dennis. 50 Elo is huge when you are already at this height.
That will be fun for you to test and find out for sure. :D
FWIW a little match on my old Q8200 at 40/1 vs. Arasan 21.4 with 6 plies start positions randomly selected (both colours) and Syzygy_5, finished
+21 -8 =11 for Schooner 2.1.
https://rwbc-chess.de

trollwatch:
Chessqueen + chessica + AlexChess + Eduard + Sylwy
D Sceviour
Posts: 570
Joined: Mon Jul 20, 2015 5:06 pm

Re: Schooner Version 2.1

Post by D Sceviour »

Guenther wrote: Wed Jul 03, 2019 4:08 pm
D Sceviour wrote: Wed Jul 03, 2019 2:16 pm
Gabor Szots wrote: Wed Jul 03, 2019 10:14 am Thank you, Dennis. 50 Elo is huge when you are already at this height.
That will be fun for you to test and find out for sure. :D
FWIW a little match on my old Q8200 at 40/1 vs. Arasan 21.4 with 6 plies start positions randomly selected (both colours) and Syzygy_5, finished
+21 -8 =11 for Schooner 2.1.
:D :D :D
User avatar
Guenther
Posts: 4605
Joined: Wed Oct 01, 2008 6:33 am
Location: Regensburg, Germany
Full name: Guenther Simon

Re: Schooner Version 2.1

Post by Guenther »

D Sceviour wrote: Wed Jul 03, 2019 4:13 pm
Guenther wrote: Wed Jul 03, 2019 4:08 pm
D Sceviour wrote: Wed Jul 03, 2019 2:16 pm
Gabor Szots wrote: Wed Jul 03, 2019 10:14 am Thank you, Dennis. 50 Elo is huge when you are already at this height.
That will be fun for you to test and find out for sure. :D
FWIW a little match on my old Q8200 at 40/1 vs. Arasan 21.4 with 6 plies start positions randomly selected (both colours) and Syzygy_5, finished
+21 -8 =11 for Schooner 2.1.
:D :D :D
BTW is it normal that it still gives a +3 eval when it is a forced exchange to available 5 men Syzygy?
(one time it played the move itself - one time opponent forced it one ply later)
Click on the dots in the game selection menu for seeing both games)

[pgn][Event "My Tournament"] [Site "RWBC Win7U-64 Q8200 2.33Ghz Nvidia GT 710"] [Date "2019.07.03"] [Round "5"] [White "Schooner_21-64"] [Black "ArasanX_214-64"] [Result "1/2-1/2"] [ECO "D00"] [Opening "Queen's pawn"] [PlyCount "95"] [Termination "adjudication"] [TimeControl "40/60"] 1. d4 {book} d5 {book} 2. Nc3 {book} Bf5 {book} 3. Bf4 {book} e6 {book} 4. e3 {+0.04/18 1.59} Bd6 {-0.07/16 2.24} 5. Bxd6 {0.00/20 3.29} Qxd6 {-0.01/17 1.64} 6. Bb5+ {+0.03/19 1.52} Nc6 {+0.01/17 1.19} 7. Nf3 {0.00/20 3.01} Ne7 {+0.04/19 3.60} 8. O-O {+0.03/19 1.88} O-O {+0.06/17 1.09} 9. Bd3 {0.00/19 2.32} a6 {+0.03/16 1.38} 10. Re1 {+0.01/20 2.45} f6 {+0.08/15 1.59} 11. Bxf5 {+0.28/20 1.19} exf5 {-0.16/16 1.70} 12. Ne2 {+0.48/17 1.89} Nd8 {-0.18/17 2.73} 13. a4 {+0.46/18 2.59} a5 {-0.18/17 1.61} 14. b3 {+0.57/18 1.28} Nec6 {-0.07/17 3.40} 15. h4 {+0.61/18 2.29} Ne6 {-0.11/18 1.55} 16. h5 {+0.54/17 1.03} f4 {-0.21/16 1.46} 17. exf4 {+0.84/19 1.18} Nxf4 {-0.35/20 1.24} 18. Nxf4 {+0.87/20 1.07} Qxf4 {-0.38/22 1.35} 19. Qe2 {+0.91/20 1.10} Rad8 {-0.43/21 1.38} 20. Qb5 {+0.82/19 1.19} Nxd4 {-0.56/22 1.55} 21. Nxd4 {+0.91/21 1.53} Qxd4 {-0.51/22 1.21} 22. Qxa5 {+0.91/23 1.23} b6 {-0.60/21 1.59} 23. Qb5 {+0.83/20 1.15} Qc3 {-0.67/20 1.59} 24. Qd3 {+0.80/22 3.37} Qxd3 {-0.76/18 2.13} 25. cxd3 {+0.80/22 0.80} Ra8 {-0.83/18 1.75} 26. Re6 {+0.97/21 1.62} Rfe8 {-0.76/19 1.25} 27. Rc6 {+1.29/22 0.92} Ra7 {-0.89/24 1.20} 28. Rac1 {+1.17/22 0.85} Re7 {-0.94/25 1.49} 29. h6 {+1.17/24 1.03} Kf7 {-0.94/22 1.24} 30. hxg7 {+1.15/25 1.08} Kxg7 {-1.14/23 1.42} 31. g4 {+1.16/25 1.22} Kg6 {-0.99/23 1.43} 32. Kg2 {+1.21/25 1.12} h6 {-1.08/21 1.42} 33. Kg3 {+1.49/21 0.97} Re2 {-1.05/22 1.44} 34. Rxc7 {+1.95/25 1.00} Rxc7 {-1.60/23 1.43} 35. Rxc7 {+2.00/27 1.95} Rb2 {-1.94/25 1.43} 36. f4 {+1.91/27 1.95} h5 {-2.25/25 1.44} 37. f5+ {+1.91/26 2.28} Kh6 {-2.37/25 1.43} 38. Rf7 {+1.87/26 1.56} hxg4 {-2.34/25 1.43} 39. Kxg4 {+2.19/27 1.59} Rxb3 {-2.57/25 1.45} 40. Rxf6+ {+2.27/28 1.65} Kg7 {-2.60/25 1.44} 41. Kg5 {+2.16/25 1.43} d4 {-2.71/23 1.71} 42. Rg6+ {+2.19/27 1.43} Kf7 {-2.38/22 1.21} 43. Kf4 {+2.11/25 1.38} Rxd3 {-2.37/24 1.22} 44. Rxb6 {+2.83/25 1.33} Rc3 {-1.74/25 2.43} 45. a5 {+2.93/30 1.53} Rc5 {0.00/23 1.16} 46. a6 {+3.23/30 1.32} d3 {0.00/26 1.03} 47. Ke3 {+3.22/32 1.30} Rxf5 {0.00/34 1.03} 48. Kxd3 {+3.24/34 1.51 - Draw by adjudication: SyzygyTB} 1/2-1/2 [Event "My Tournament"] [Site "RWBC Win7U-64 Q8200 2.33Ghz Nvidia GT 710"] [Date "2019.07.03"] [Round "9"] [White "ArasanX_214-64"] [Black "Schooner_21-64"] [Result "1/2-1/2"] [ECO "A80"] [Opening "Dutch - 2.Bg5 Variation"] [PlyCount "129"] [Termination "adjudication"] [TimeControl "40/60"] 1. d4 {book} f5 {book} 2. Bg5 {book} h6 {book} 3. Bh4 {book} g5 {book} 4. e3 {+0.46/18 1.85} Nf6 {-0.32/17 1.49} 5. Bg3 {+0.60/19 1.71} Bg7 {-0.39/18 1.58} 6. Bd3 {+0.56/17 3.14} d5 {-0.36/18 2.02} 7. h4 {+0.67/16 2.54} g4 {-0.23/18 1.89} 8. Ne2 {+0.73/16 2.17} c5 {-0.23/18 1.66} 9. dxc5 {+0.75/16 1.65} Nh5 {0.00/19 1.71} 10. c3 {+0.56/17 1.52} O-O {+0.21/19 1.57} 11. Nd2 {+0.51/16 2.09} a5 {+0.28/17 4.09} 12. Bxb8 {+0.57/17 2.00} Rxb8 {+0.10/20 2.57} 13. Qb3 {+0.69/18 1.22} e5 {+0.01/17 1.02} 14. O-O-O {+0.58/17 1.73} Be6 {+0.21/19 2.13} 15. f4 {+0.61/17 1.38} gxf3 {-0.08/18 3.09} 16. gxf3 {+0.85/18 1.27} Qc8 {0.00/18 0.88} 17. Qb6 {+1.42/17 1.26} Rf6 {-0.14/20 2.45} 18. Rhg1 {+1.50/18 2.42} e4 {-0.04/18 1.38} 19. Bc2 {+1.32/18 2.22} Bf7 {-0.13/18 0.87} 20. Qb5 {+1.28/20 1.62} b6 {-0.27/19 1.95} 21. c6 {+1.43/19 1.24} Qxc6 {-0.42/20 2.47} 22. Qxc6 {+1.75/21 1.03} Rxc6 {-0.41/20 3.38} 23. Nd4 {+1.57/20 1.44} Rf6 {-0.41/22 0.96} 24. fxe4 {+1.22/18 3.59} fxe4 {-0.35/21 2.12} 25. c4 {+0.72/18 1.56} Rf2 {-0.35/21 0.92} 26. cxd5 {+1.10/20 2.07} Bxd5 {-0.44/21 0.68} 27. Nb1 {+0.77/17 1.38} Nf6 {-0.54/20 3.68} 28. Nc3 {+0.61/18 1.41} Kh8 {-0.57/22 0.94} 29. a3 {+0.75/19 1.94} Re8 {-0.48/20 0.80} 30. Ba4 {+0.52/18 1.80} Re5 {-0.51/19 0.71} 31. Nde2 {+0.96/18 1.59} Bf7 {-0.54/17 1.15} 32. Nf4 {+1.03/17 1.02} Rf3 {-0.79/19 2.12} 33. Rd8+ {+0.65/17 1.00} Kh7 {-0.25/21 0.64} 34. Re1 {+0.75/18 1.01} Rf2 {-0.57/16 0.90} 35. Rb8 {+1.18/16 1.02} Nh5 {-0.56/17 0.69} 36. Bc2 {+1.01/17 1.01} Ng3 {-0.56/18 1.58} 37. Rb7 {+0.60/17 1.02} Kg8 {-0.56/19 0.87} 38. Kb1 {+0.57/17 1.01} Nf5 {-0.52/18 0.94} 39. Rb8+ {+0.56/16 1.02} Re8 {-0.22/17 0.62} 40. Rxe8+ {+0.53/18 1.02} Bxe8 {-0.29/20 0.94} 41. Nxe4 {+0.44/19 2.53} Rh2 {-0.10/21 1.49} 42. Nd5 {+0.28/20 2.30} Bc6 {-0.17/23 2.01} 43. Nec3 {+0.94/20 1.05} Bxc3 {-0.13/24 2.30} 44. Nxc3 {+0.44/23 1.68} Nxh4 {-0.15/23 1.45} 45. e4 {+0.39/21 2.08} Kg7 {0.00/21 2.82} 46. Rd1 {+0.35/21 1.17} Rf2 {0.00/22 2.56} 47. Kc1 {+0.12/19 2.62} Ng6 {+0.46/19 1.24} 48. Rg1 {-0.03/18 1.40} h5 {+0.67/21 1.12} 49. Nd5 {-0.72/20 3.23} h4 {+1.03/23 1.15} 50. Rh1 {-0.69/22 1.56} Bd7 {+1.12/22 1.20} 51. Nxb6 {-0.96/21 2.22} Be6 {+1.18/24 1.19} 52. b4 {-0.69/21 1.24} axb4 {+1.26/25 1.15} 53. axb4 {-0.57/21 0.95} h3 {+1.49/25 1.17} 54. Nd5 {-0.80/22 1.39} h2 {+1.64/23 1.30} 55. b5 {-0.75/22 2.03} Bg4 {+3.15/21 1.27} 56. b6 {-0.12/22 0.90} Bf3 {+2.99/24 0.99} 57. b7 {0.00/25 0.91} Bxh1 {+3.01/27 1.02} 58. b8=Q {0.00/27 0.92} Rxc2+ {+3.01/27 1.33} 59. Kxc2 {0.00/28 0.95} Bxe4+ {+3.14/25 2.47} 60. Kd2 {0.00/28 0.94} h1=Q {+3.33/19 0.93} 61. Qa7+ {0.00/29 0.97} Kf8 {+3.42/22 1.91} 62. Qb8+ {0.00/29 0.99} Kf7 {+3.46/22 1.23} 63. Qc7+ {0.00/30 1.00} Ke6 {+3.44/22 1.10} 64. Nf4+ {0.00/31 1.01} Nxf4 {+3.38/21 1.18} 65. Qxf4 {0.00/44 1.02 - Draw by adjudication: SyzygyTB} 1/2-1/2[/pgn]
https://rwbc-chess.de

trollwatch:
Chessqueen + chessica + AlexChess + Eduard + Sylwy
D Sceviour
Posts: 570
Joined: Mon Jul 20, 2015 5:06 pm

Re: Schooner Version 2.1

Post by D Sceviour »

Guenther wrote: Wed Jul 03, 2019 7:04 pm BTW is it normal that it still gives a +3 eval when it is a forced exchange to available 5 men Syzygy?
(one time it played the move itself - one time opponent forced it one ply later)
Yes, this is normal. In the first game, the engine calculates for six pieces (not five) and scores accordingly with the move:

[d]8/5k2/PR6/5r2/8/3pK3/8/8 w - - 0 2

Then the GUI adjudicates using its 5-piece egtb. If the GUI uses adjudication, then the engine egtb has no effect. Syzygy support was mainly included for ICS tournaments, or where users do not adjudicate games. It can also be useful for endgame analysis.
Max
Posts: 247
Joined: Tue Apr 13, 2010 10:41 am

Re: Schooner Version 2.1

Post by Max »

D Sceviour wrote: Wed Jul 03, 2019 2:15 pm
Max wrote: Wed Jul 03, 2019 12:00 pm Unfortunately Schooner 2.1 crashed running with wine 4 on macOS.
Is there any hope for a (native) mac version?
Sorry, Schooner only runs on windows. I would like to create a Linux build. However, it seems like there is always one more thing to do. Day after week after month passes and an update never gets released. Are there any Windows->Mac converters besides Wine you might try?
No, there is not such a converter. We can use a VM (like VirtualBox) with a full blown Windows installation plus chess gui. But this would be an overkill.

Many (or most, if they don't use AVX2) windows engines run with wine, but Schooner unfortunately not. Too sad.
Hope we're not just the biological boot loader for digital super intelligence. Unfortunately, that is increasingly probable - Elon Musk