Gaviota not working with Rybka Aquarium

[mhalstern]

Miguel

I have Gaviota working with Fritz12 with a standard wb2uci setup. However, as a normal winboard engine, it does not work in Rybka Aquarium. It loads into ram and takes up cpu usage, as if it is playing, however it simply always loses on time, without having made any moves.

It is behaving as if infinite analysis has been called (set timer to infinite) instead of recognizing the time control.

Is this a known issue?

Here is gaviota.ini.txt:

;=========================
; parameters for Gaviota
;=========================

;-------------------------------------------------------
; name that Winboard will use for saving games into pgn
;-------------------------------------------------------

name = "Gaviota 0.74.12"

;------------------------------------------------------------
; hashmem --> is hash tables (transposition, refutation).
; ref_table --> is a special table for refutation moves,
; not clear what the best value is, but
; I would try double the hashmem or the same.
;
; learn_table --> holds information from learn.epd
; do not set it bigger than 1 M
; pawn_table --> information about pawn structure.
; keep it small, 1-4 M
; M is Megabytes, k kilobytes , b bytes
;-------------------------------------------------------------

hashmem = 1024 M
ref_table = 32 M
learn_table = 1 M
pawn_table = 1 M

;----------------------------------------------
; log info. max_logs is number of files kept.
; After that number is reached, it will start
; to overwrite them
;----------------------------------------------

log = on
max_logs = 20
log_path = "logs"

;-----------------------------------------------------------------------
; ponder can have three values, 0, 1, or 2 (recommended).
; 2: Pondering will be determined by the interface (winboard, etc.)
; 1: Ponder is on, hardwired, no matter what the interface wants
; 0: Ponder is off, hardwired, won't turn on even if the interface wants
;-----------------------------------------------------------------------

ponder = 0

;-------------
; book info
;-------------

book = off
book_file = "book2650.bin"
book_learn = on

;------------------------------------------------
; learning of positions when is out of book.
; learn = on --> will save what just learned,
; use_learned = on --> will use any previous stored
; information in the "learn_file"
; learn_file = "/properpath/whatevername.epd"
;------------------------------------------------

learn = on
use_learned = on
learn_file = "learn.epd"

;-------------------------------------------------------------------
; won't show every output detail in the first n plies of the search
;-------------------------------------------------------------------

quiet_level = 6

;---------------------------------------------------------------------------
; logs when the node/sec is lower than "alert_nps"
; it is useful to detect when the program slowed down because of the OS
; hardware problems, paging, etc.
;---------------------------------------------------------------------------

alert_nps = 10000

;-----------------------------------------------------------------
; Runs all the commands included in the file after initialization
; and before accepting commands from the user.
; Delete the command if you do not want to have an initial script
; or comment it out placing a semicolon before.
;-----------------------------------------------------------------

;script_file = "input.txt"

;-----------------------------
; Resign in hopeless position?
;-----------------------------

resign_mode = on

;------------------------------------------------------
; Should I kibitz info on computer tournaments online?
;------------------------------------------------------

kibitz = off

;---------------------------------
;
; Gaviota Tablebase parameters
;
;---------------------------------

use_tablebase = on
tb_probing_depth = 2
tb_pieces_used = 5
tb_decoding_scheme = 4
tb_compression = on
tb_cache = on
tb_cache_size = 64 M
tb_cache_slot_size = 8 k

tb_path_1 = "gtb/"
tb_path_2 = ""
tb_path_3 = ""
tb_path_4 = ""

;-------------------------------------------------------------------
; SAN support (Standard Algebraic Notation) (e.g. Nf3)
; on ==> will try to send and accept SAN moves if GUI supports it
; off ==> will use default coordinate notation (e.g. g1f3)
;
; Some GUIs (Arena, ChessGUI etc.) are NOT fully SAN compliant
; and send wrong SAN moves. Use SAN = off with those GUIs
; I believe it is pretty safe to leave this off
;-------------------------------------------------------------------

SAN = off

;------------------------------------------------------------
; PARALLEL SEARCH
; cores --> CPU Usage
; Gaviota will launch up a number of threads
; equal to this value
; Supported up to 16 cores
;------------------------------------------------------------

cores = 4

;-------
; end!
;-------





Here are some log files showing the output:


Gaviota 0.74.12
Compiled: Jan 5 2010 23:42:50

Playing Session
---------------


learning succesfully loaded from: learn.epd

SEED for rand_table_init(): 1274486350

received <-- xboard
received <-- easy
received <-- post
received <-- new


NEW_GAME


SEED for rand_table_init(): 1274486350

received <-- post
received <-- force
received <-- white
Game Time: 0.000000
white --> move received
Game Time: 0.000000
move received was NOT identified
received <-- analyze
About to launch an engine thread...
About to create a thread...
received <-- exit
Going to stop the engine...
Decision Phase...

-------------- NEW ITERATION ----------------

Signature[1]: 530f7644
rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQkq - 0 1
+-----------------+
| r n b q k b n r |
| p p p p p p p p |
| . . . . . . . . |
| . . . . . . . . | Castling: KQkq
| . . . . . . . . | ep: -
| . . . . . . . . |
| P P P P P P P P |
| R N B Q K B N R | [White]
+-----------------+

legal moves available = 20
deepening call...
Creating thread...
t=0
thread created...
iterative deepening wrapper...
********* Starts iterative deepening, thread = 0
set timer to infinite
1 1 0.0 -0.09 1.f4
2 1 0.0 +0.15 1.c4
3 1 0.0 +0.66 1.e4
11 1 0.0 +0.80 1.e3
~ +0.39 will fail low? ( 96)
22 2 0.0 1.e3
89 2 0.0 +0.00 1.e3 e6
261 3 0.0 +0.65 1.e3 Nf6 2.Nc3
~ +0.25 will fail low? ( 60)
405 4 0.0 1.e3
1267 4 0.0 -0.00 1.e3 Nf6 2.Nc3 d5
1928 4 0.0 +0.00 1.d4 d5 2.Qd3 Qd6
7192 5 0.0 +0.48 1.d4 Nf6 2.Qd3 d5 3.Nf3
~ +0.07 will fail low? ( 17)
12232 6 0.0 1.d4
28134 6 0.0
36258 6 0.0 +0.02 1.d4 Nf6 2.Qd3 d5 3.Qb5+ Nc6 4.Nf3
56545 6 0.0 +0.02 1.Nc3 Nf6 2.e4 d5 3.Qf3 c6 4.exd5 cxd5
56768 6: 0.0 +0.02 1.Nc3 Nf6 2.e4 d5 3.Qf3 c6 4.exd5 cxd5
81723 7 0.0 1.Nc3
124385 7 0.1 +0.52 1.Nc3 Nc6 2.Nf3 d5 3.d4 Qd6 4.Bg5
156445 7: 0.1 +0.52 1.Nc3 Nc6 2.Nf3 d5 3.d4 Qd6 4.Bg5
~ +0.11 will fail low? ( 26)
179185 8 0.1 1.Nc3
379705 8 0.2
429976 8 0.2 +0.00 1.Nc3 Nc6 2.Nf3 Nf6 3.d3 d6 4.Be3 Be6
682411 8: 0.3 +0.00 1.Nc3 Nc6 2.Nf3 Nf6 3.d3 d6 4.Be3 Be6
932779 9 0.4 +0.25 1.Nc3 Nc6 2.d4 Nf6 3.Bf4 d6 4.e3 Bf5
5.Bc4
1570031 9 0.6 +0.34 1.d4 Nf6 2.Qd3 Nc6 3.Nf3 e6 4.a3 Bd6
5.g3
1666764 9: 0.6 +0.34 1.d4 Nf6 2.Qd3 Nc6 3.Nf3 e6 4.a3 Bd6
5.g3
3096787 10 1.1 +0.14 1.d4 Nf6 2.Nc3 d6 3.e4 c6 4.e5 dxe5
5.dxe5 Qxd1+ 6.Kxd1 Nd5 7.Nxd5 cxd5
5121292 10: 1.7 +0.14 1.d4 Nf6 2.Nc3 d6 3.e4 c6 4.e5 dxe5
5.dxe5 Qxd1+ 6.Kxd1 Nd5 7.Nxd5 cxd5
14825093 11 5.1 +0.39 1.d4 Nf6 2.Nc3 h6 3.e4 d5 4.exd5 Nxd5
5.Qf3 Nb4 6.Bb5+ N8c6
15393491 11: 5.3 +0.39 1.d4 Nf6 2.Nc3 h6 3.e4 d5 4.exd5 Nxd5
5.Qf3 Nb4 6.Bb5+ N8c6
~ -0.02 will fail low? ( -4)
~ +0.08 ok ( 20)
~ -0.02 will fail low? ( -4)
~ +0.38 ok ( 94)
35066526 12 12.1 +0.20 1.d4 Nf6 2.Nc3 d5 3.Bf4 Bf5 4.h3 h6
5.g4 g5 6.Bxc7 Qxc7 7.gxf5
52650509 12: 17.4 +0.20 1.d4 Nf6 2.Nc3 d5 3.Bf4 Bf5 4.h3 h6
5.g4 g5 6.Bxc7 Qxc7 7.gxf5
108716286 13 36.1 +0.27 1.d4 Nf6 2.Nc3 h6 3.Nf3 d5 4.Qd3 Nc6
5.Ne5 Qd6 6.Nb5 Qb4+ 7.Bd2 Nxe5 8.Bxb4
Nxd3+ 9.cxd3
117517266 13: 39.5 +0.27 1.d4 Nf6 2.Nc3 h6 3.Nf3 d5 4.Qd3 Nc6
5.Ne5 Qd6 6.Nb5 Qb4+ 7.Bd2 Nxe5 8.Bxb4
Nxd3+ 9.cxd3


Gaviota 0.74.12
Compiled: Jan 5 2010 23:42:50

Playing Session
---------------


learning succesfully loaded from: learn.epd

SEED for rand_table_init(): 1274486802

received <-- xboard
received <-- easy
received <-- post
received <-- new


NEW_GAME


SEED for rand_table_init(): 1274486802

received <-- post
received <-- force
received <-- white
Game Time: 0.000000
white --> move received
Game Time: 0.000000
move received was NOT identified
received <-- analyze
About to launch an engine thread...
About to create a thread...
received <-- exit
Going to stop the engine...
Decision Phase...

-------------- NEW ITERATION ----------------

Signature[1]: 530f7644
rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQkq - 0 1
+-----------------+
| r n b q k b n r |
| p p p p p p p p |
| . . . . . . . . |
| . . . . . . . . | Castling: KQkq
| . . . . . . . . | ep: -
| . . . . . . . . |
| P P P P P P P P |
| R N B Q K B N R | [White]
+-----------------+

legal moves available = 20
deepening call...
Creating thread...
t=0
thread created...
iterative deepening wrapper...
********* Starts iterative deepening, thread = 0
set timer to infinite
1 1 0.0 -0.09 1.f4
2 1 0.0 +0.15 1.c4
3 1 0.0 +0.66 1.e4
11 1 0.0 +0.80 1.e3
~ +0.39 will fail low? ( 96)
22 2 0.0 1.e3
89 2 0.0 +0.00 1.e3 e6
261 3 0.0 +0.65 1.e3 Nf6 2.Nc3
~ +0.25 will fail low? ( 60)
405 4 0.0 1.e3
1259 4 0.0 -0.00 1.e3 Nf6 2.Nc3 d5
1931 4 0.0 +0.00 1.d4 d5 2.Qd3 Qd6
7097 5 0.0 +0.48 1.d4 Nf6 2.Qd3 d5 3.Nf3
~ +0.07 will fail low? ( 17)
13386 6 0.0 1.d4
29399 6 0.0
37534 6 0.0 +0.02 1.d4 Nf6 2.Qd3 d5 3.Qb5+ Nc6 4.Nf3
56571 6 0.0 +0.02 1.Nc3 Nf6 2.e4 d5 3.Qf3 c6 4.exd5 cxd5
56898 6: 0.0 +0.02 1.Nc3 Nf6 2.e4 d5 3.Qf3 c6 4.exd5 cxd5
85143 7 0.0 1.Nc3
130227 7 0.1 +0.52 1.Nc3 Nc6 2.Nf3 d5 3.d4 Qd6 4.Bg5
154040 7: 0.1 +0.52 1.Nc3 Nc6 2.Nf3 d5 3.d4 Qd6 4.Bg5
~ +0.11 will fail low? ( 26)
176429 8 0.1 1.Nc3
389424 8 0.2
441411 8 0.2 +0.00 1.Nc3 Nc6 2.Nf3 Nf6 3.d3 d6 4.Be3 Be6
689973 8: 0.3 +0.00 1.Nc3 Nc6 2.Nf3 Nf6 3.d3 d6 4.Be3 Be6
981817 9 0.4 +0.25 1.Nc3 Nc6 2.d4 Nf6 3.Bf4 d6 4.e3 Bf5
5.Bc4
1589876 9 0.6 +0.25 1.d4 Nf6 2.Qd3 c6 3.Nf3 d6 4.a4 Be6
5.a5
1843889 9: 0.6 +0.25 1.d4 Nf6 2.Qd3 c6 3.Nf3 d6 4.a4 Be6
5.a5
3346181 10 1.2 +0.14 1.d4 Nf6 2.Nc3 d6 3.e4 c6 4.e5 dxe5
5.dxe5 Qxd1+ 6.Kxd1 Nd5 7.Nxd5 cxd5
5426523 10: 1.8 +0.14 1.d4 Nf6 2.Nc3 d6 3.e4 c6 4.e5 dxe5
5.dxe5 Qxd1+ 6.Kxd1 Nd5 7.Nxd5 cxd5
13918424 11 4.8 +0.39 1.d4 Nf6 2.Nc3 h6 3.e4 d5 4.exd5 Nxd5
5.Qf3 Nb4 6.Bb5+ N8c6
14498624 11: 5.0 +0.39 1.d4 Nf6 2.Nc3 h6 3.e4 d5 4.exd5 Nxd5
5.Qf3 Nb4 6.Bb5+ N8c6
~ -0.02 will fail low? ( -4)
~ +0.08 ok ( 20)
~ -0.02 will fail low? ( -4)
~ +0.80 ok ( 196)
34001335 12 11.6 +0.18 1.d4 d5 2.Bf4 Bf5 3.e3 Nf6 4.Bd3 Bxd3
5.cxd3 e6 6.Qb3 b6 7.Qa4+ Qd7 8.Qxd7+
Kxd7
54345487 12: 17.8 +0.18 1.d4 d5 2.Bf4 Bf5 3.e3 Nf6 4.Bd3 Bxd3
5.cxd3 e6 6.Qb3 b6 7.Qa4+ Qd7 8.Qxd7+
Kxd7
112487237 13 36.5 +0.27 1.d4 Nf6 2.Nc3 h6 3.Nf3 d5 4.Ne5 Nc6
5.Qd3 Qd6 6.Nb5 Qb4+ 7.Bd2 Nxe5 8.Bxb4
Nxd3+ 9.cxd3
133593166 13: 45.3 +0.27 1.d4 Nf6 2.Nc3 h6 3.Nf3 d5 4.Ne5 Nc6
5.Qd3 Qd6 6.Nb5 Qb4+ 7.Bd2 Nxe5 8.Bxb4
Nxd3+ 9.cxd3


Thanks

[michiguel]

Miguel

I have Gaviota working with Fritz12 with a standard wb2uci setup. However, as a normal winboard engine, it does not work in Rybka Aquarium. It loads into ram and takes up cpu usage, as if it is playing, however it simply always loses on time, without having made any moves.

It is behaving as if infinite analysis has been called (set timer to infinite) instead of recognizing the time control.

Is this a known issue?

Yes, unfortunately there is very little I can do, realistically. I am sorry to say that since you came with a perfect log and excellent bug report.

I tried several things and we emailed back and forth with Carlos Pesce (who reported it to me) and we could not solve it. Whatever Aquarium is doing, is not standard. I even wonder whether there is any WB engine that works.
If you look at the logs, Aq. is trying to get Gaviota to play making it to "analyze". That is crazy. It may be trying to use the UCI spirit through WB commands.

To make things even worse, Gaviota works with the demo but not with the "paid" version. So, whatever I do with the demo is useless and I am not going to buy the product just to fix their problems I am fairly confident that if I have it in my hands, I would figure it out; but...

...long story short, I am writing a UCI version of Gaviota (I am almost done, I am trying to add multiPV since I am at it). Honestly, it was easier than trying to deal with a GUI that is sabotaging a protocol. At the end, it will be better to have both options anyway.

Miguel

Here is gaviota.ini.txt:

;=========================
; parameters for Gaviota
;=========================

;-------------------------------------------------------
; name that Winboard will use for saving games into pgn
;-------------------------------------------------------

name = "Gaviota 0.74.12"

;------------------------------------------------------------
; hashmem --> is hash tables (transposition, refutation).
; ref_table --> is a special table for refutation moves,
; not clear what the best value is, but
; I would try double the hashmem or the same.
;
; learn_table --> holds information from learn.epd
; do not set it bigger than 1 M
; pawn_table --> information about pawn structure.
; keep it small, 1-4 M
; M is Megabytes, k kilobytes , b bytes
;-------------------------------------------------------------

hashmem = 1024 M
ref_table = 32 M
learn_table = 1 M
pawn_table = 1 M

;----------------------------------------------
; log info. max_logs is number of files kept.
; After that number is reached, it will start
; to overwrite them
;----------------------------------------------

log = on
max_logs = 20
log_path = "logs"

;-----------------------------------------------------------------------
; ponder can have three values, 0, 1, or 2 (recommended).
; 2: Pondering will be determined by the interface (winboard, etc.)
; 1: Ponder is on, hardwired, no matter what the interface wants
; 0: Ponder is off, hardwired, won't turn on even if the interface wants
;-----------------------------------------------------------------------

ponder = 0

;-------------
; book info
;-------------

book = off
book_file = "book2650.bin"
book_learn = on

;------------------------------------------------
; learning of positions when is out of book.
; learn = on --> will save what just learned,
; use_learned = on --> will use any previous stored
; information in the "learn_file"
; learn_file = "/properpath/whatevername.epd"
;------------------------------------------------

learn = on
use_learned = on
learn_file = "learn.epd"

;-------------------------------------------------------------------
; won't show every output detail in the first n plies of the search
;-------------------------------------------------------------------

quiet_level = 6

;---------------------------------------------------------------------------
; logs when the node/sec is lower than "alert_nps"
; it is useful to detect when the program slowed down because of the OS
; hardware problems, paging, etc.
;---------------------------------------------------------------------------

alert_nps = 10000

;-----------------------------------------------------------------
; Runs all the commands included in the file after initialization
; and before accepting commands from the user.
; Delete the command if you do not want to have an initial script
; or comment it out placing a semicolon before.
;-----------------------------------------------------------------

;script_file = "input.txt"

;-----------------------------
; Resign in hopeless position?
;-----------------------------

resign_mode = on

;------------------------------------------------------
; Should I kibitz info on computer tournaments online?
;------------------------------------------------------

kibitz = off

;---------------------------------
;
; Gaviota Tablebase parameters
;
;---------------------------------

use_tablebase = on
tb_probing_depth = 2
tb_pieces_used = 5
tb_decoding_scheme = 4
tb_compression = on
tb_cache = on
tb_cache_size = 64 M
tb_cache_slot_size = 8 k

tb_path_1 = "gtb/"
tb_path_2 = ""
tb_path_3 = ""
tb_path_4 = ""

;-------------------------------------------------------------------
; SAN support (Standard Algebraic Notation) (e.g. Nf3)
; on ==> will try to send and accept SAN moves if GUI supports it
; off ==> will use default coordinate notation (e.g. g1f3)
;
; Some GUIs (Arena, ChessGUI etc.) are NOT fully SAN compliant
; and send wrong SAN moves. Use SAN = off with those GUIs
; I believe it is pretty safe to leave this off
;-------------------------------------------------------------------

SAN = off

;------------------------------------------------------------
; PARALLEL SEARCH
; cores --> CPU Usage
; Gaviota will launch up a number of threads
; equal to this value
; Supported up to 16 cores
;------------------------------------------------------------

cores = 4

;-------
; end!
;-------





Here are some log files showing the output:


Gaviota 0.74.12
Compiled: Jan 5 2010 23:42:50

Playing Session
---------------


learning succesfully loaded from: learn.epd

SEED for rand_table_init(): 1274486350

received <-- xboard
received <-- easy
received <-- post
received <-- new


NEW_GAME


SEED for rand_table_init(): 1274486350

received <-- post
received <-- force
received <-- white
Game Time: 0.000000
white --> move received
Game Time: 0.000000
move received was NOT identified
received <-- analyze
About to launch an engine thread...
About to create a thread...
received <-- exit
Going to stop the engine...
Decision Phase...

-------------- NEW ITERATION ----------------

Signature[1]: 530f7644
rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQkq - 0 1
+-----------------+
| r n b q k b n r |
| p p p p p p p p |
| . . . . . . . . |
| . . . . . . . . | Castling: KQkq
| . . . . . . . . | ep: -
| . . . . . . . . |
| P P P P P P P P |
| R N B Q K B N R | [White]
+-----------------+

legal moves available = 20
deepening call...
Creating thread...
t=0
thread created...
iterative deepening wrapper...
********* Starts iterative deepening, thread = 0
set timer to infinite
1 1 0.0 -0.09 1.f4
2 1 0.0 +0.15 1.c4
3 1 0.0 +0.66 1.e4
11 1 0.0 +0.80 1.e3
~ +0.39 will fail low? ( 96)
22 2 0.0 1.e3
89 2 0.0 +0.00 1.e3 e6
261 3 0.0 +0.65 1.e3 Nf6 2.Nc3
~ +0.25 will fail low? ( 60)
405 4 0.0 1.e3
1267 4 0.0 -0.00 1.e3 Nf6 2.Nc3 d5
1928 4 0.0 +0.00 1.d4 d5 2.Qd3 Qd6
7192 5 0.0 +0.48 1.d4 Nf6 2.Qd3 d5 3.Nf3
~ +0.07 will fail low? ( 17)
12232 6 0.0 1.d4
28134 6 0.0
36258 6 0.0 +0.02 1.d4 Nf6 2.Qd3 d5 3.Qb5+ Nc6 4.Nf3
56545 6 0.0 +0.02 1.Nc3 Nf6 2.e4 d5 3.Qf3 c6 4.exd5 cxd5
56768 6: 0.0 +0.02 1.Nc3 Nf6 2.e4 d5 3.Qf3 c6 4.exd5 cxd5
81723 7 0.0 1.Nc3
124385 7 0.1 +0.52 1.Nc3 Nc6 2.Nf3 d5 3.d4 Qd6 4.Bg5
156445 7: 0.1 +0.52 1.Nc3 Nc6 2.Nf3 d5 3.d4 Qd6 4.Bg5
~ +0.11 will fail low? ( 26)
179185 8 0.1 1.Nc3
379705 8 0.2
429976 8 0.2 +0.00 1.Nc3 Nc6 2.Nf3 Nf6 3.d3 d6 4.Be3 Be6
682411 8: 0.3 +0.00 1.Nc3 Nc6 2.Nf3 Nf6 3.d3 d6 4.Be3 Be6
932779 9 0.4 +0.25 1.Nc3 Nc6 2.d4 Nf6 3.Bf4 d6 4.e3 Bf5
5.Bc4
1570031 9 0.6 +0.34 1.d4 Nf6 2.Qd3 Nc6 3.Nf3 e6 4.a3 Bd6
5.g3
1666764 9: 0.6 +0.34 1.d4 Nf6 2.Qd3 Nc6 3.Nf3 e6 4.a3 Bd6
5.g3
3096787 10 1.1 +0.14 1.d4 Nf6 2.Nc3 d6 3.e4 c6 4.e5 dxe5
5.dxe5 Qxd1+ 6.Kxd1 Nd5 7.Nxd5 cxd5
5121292 10: 1.7 +0.14 1.d4 Nf6 2.Nc3 d6 3.e4 c6 4.e5 dxe5
5.dxe5 Qxd1+ 6.Kxd1 Nd5 7.Nxd5 cxd5
14825093 11 5.1 +0.39 1.d4 Nf6 2.Nc3 h6 3.e4 d5 4.exd5 Nxd5
5.Qf3 Nb4 6.Bb5+ N8c6
15393491 11: 5.3 +0.39 1.d4 Nf6 2.Nc3 h6 3.e4 d5 4.exd5 Nxd5
5.Qf3 Nb4 6.Bb5+ N8c6
~ -0.02 will fail low? ( -4)
~ +0.08 ok ( 20)
~ -0.02 will fail low? ( -4)
~ +0.38 ok ( 94)
35066526 12 12.1 +0.20 1.d4 Nf6 2.Nc3 d5 3.Bf4 Bf5 4.h3 h6
5.g4 g5 6.Bxc7 Qxc7 7.gxf5
52650509 12: 17.4 +0.20 1.d4 Nf6 2.Nc3 d5 3.Bf4 Bf5 4.h3 h6
5.g4 g5 6.Bxc7 Qxc7 7.gxf5
108716286 13 36.1 +0.27 1.d4 Nf6 2.Nc3 h6 3.Nf3 d5 4.Qd3 Nc6
5.Ne5 Qd6 6.Nb5 Qb4+ 7.Bd2 Nxe5 8.Bxb4
Nxd3+ 9.cxd3
117517266 13: 39.5 +0.27 1.d4 Nf6 2.Nc3 h6 3.Nf3 d5 4.Qd3 Nc6
5.Ne5 Qd6 6.Nb5 Qb4+ 7.Bd2 Nxe5 8.Bxb4
Nxd3+ 9.cxd3


Gaviota 0.74.12
Compiled: Jan 5 2010 23:42:50

Playing Session
---------------


learning succesfully loaded from: learn.epd

SEED for rand_table_init(): 1274486802

received <-- xboard
received <-- easy
received <-- post
received <-- new


NEW_GAME


SEED for rand_table_init(): 1274486802

received <-- post
received <-- force
received <-- white
Game Time: 0.000000
white --> move received
Game Time: 0.000000
move received was NOT identified
received <-- analyze
About to launch an engine thread...
About to create a thread...
received <-- exit
Going to stop the engine...
Decision Phase...

-------------- NEW ITERATION ----------------

Signature[1]: 530f7644
rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQkq - 0 1
+-----------------+
| r n b q k b n r |
| p p p p p p p p |
| . . . . . . . . |
| . . . . . . . . | Castling: KQkq
| . . . . . . . . | ep: -
| . . . . . . . . |
| P P P P P P P P |
| R N B Q K B N R | [White]
+-----------------+

legal moves available = 20
deepening call...
Creating thread...
t=0
thread created...
iterative deepening wrapper...
********* Starts iterative deepening, thread = 0
set timer to infinite
1 1 0.0 -0.09 1.f4
2 1 0.0 +0.15 1.c4
3 1 0.0 +0.66 1.e4
11 1 0.0 +0.80 1.e3
~ +0.39 will fail low? ( 96)
22 2 0.0 1.e3
89 2 0.0 +0.00 1.e3 e6
261 3 0.0 +0.65 1.e3 Nf6 2.Nc3
~ +0.25 will fail low? ( 60)
405 4 0.0 1.e3
1259 4 0.0 -0.00 1.e3 Nf6 2.Nc3 d5
1931 4 0.0 +0.00 1.d4 d5 2.Qd3 Qd6
7097 5 0.0 +0.48 1.d4 Nf6 2.Qd3 d5 3.Nf3
~ +0.07 will fail low? ( 17)
13386 6 0.0 1.d4
29399 6 0.0
37534 6 0.0 +0.02 1.d4 Nf6 2.Qd3 d5 3.Qb5+ Nc6 4.Nf3
56571 6 0.0 +0.02 1.Nc3 Nf6 2.e4 d5 3.Qf3 c6 4.exd5 cxd5
56898 6: 0.0 +0.02 1.Nc3 Nf6 2.e4 d5 3.Qf3 c6 4.exd5 cxd5
85143 7 0.0 1.Nc3
130227 7 0.1 +0.52 1.Nc3 Nc6 2.Nf3 d5 3.d4 Qd6 4.Bg5
154040 7: 0.1 +0.52 1.Nc3 Nc6 2.Nf3 d5 3.d4 Qd6 4.Bg5
~ +0.11 will fail low? ( 26)
176429 8 0.1 1.Nc3
389424 8 0.2
441411 8 0.2 +0.00 1.Nc3 Nc6 2.Nf3 Nf6 3.d3 d6 4.Be3 Be6
689973 8: 0.3 +0.00 1.Nc3 Nc6 2.Nf3 Nf6 3.d3 d6 4.Be3 Be6
981817 9 0.4 +0.25 1.Nc3 Nc6 2.d4 Nf6 3.Bf4 d6 4.e3 Bf5
5.Bc4
1589876 9 0.6 +0.25 1.d4 Nf6 2.Qd3 c6 3.Nf3 d6 4.a4 Be6
5.a5
1843889 9: 0.6 +0.25 1.d4 Nf6 2.Qd3 c6 3.Nf3 d6 4.a4 Be6
5.a5
3346181 10 1.2 +0.14 1.d4 Nf6 2.Nc3 d6 3.e4 c6 4.e5 dxe5
5.dxe5 Qxd1+ 6.Kxd1 Nd5 7.Nxd5 cxd5
5426523 10: 1.8 +0.14 1.d4 Nf6 2.Nc3 d6 3.e4 c6 4.e5 dxe5
5.dxe5 Qxd1+ 6.Kxd1 Nd5 7.Nxd5 cxd5
13918424 11 4.8 +0.39 1.d4 Nf6 2.Nc3 h6 3.e4 d5 4.exd5 Nxd5
5.Qf3 Nb4 6.Bb5+ N8c6
14498624 11: 5.0 +0.39 1.d4 Nf6 2.Nc3 h6 3.e4 d5 4.exd5 Nxd5
5.Qf3 Nb4 6.Bb5+ N8c6
~ -0.02 will fail low? ( -4)
~ +0.08 ok ( 20)
~ -0.02 will fail low? ( -4)
~ +0.80 ok ( 196)
34001335 12 11.6 +0.18 1.d4 d5 2.Bf4 Bf5 3.e3 Nf6 4.Bd3 Bxd3
5.cxd3 e6 6.Qb3 b6 7.Qa4+ Qd7 8.Qxd7+
Kxd7
54345487 12: 17.8 +0.18 1.d4 d5 2.Bf4 Bf5 3.e3 Nf6 4.Bd3 Bxd3
5.cxd3 e6 6.Qb3 b6 7.Qa4+ Qd7 8.Qxd7+
Kxd7
112487237 13 36.5 +0.27 1.d4 Nf6 2.Nc3 h6 3.Nf3 d5 4.Ne5 Nc6
5.Qd3 Qd6 6.Nb5 Qb4+ 7.Bd2 Nxe5 8.Bxb4
Nxd3+ 9.cxd3
133593166 13: 45.3 +0.27 1.d4 Nf6 2.Nc3 h6 3.Nf3 d5 4.Ne5 Nc6
5.Qd3 Qd6 6.Nb5 Qb4+ 7.Bd2 Nxe5 8.Bxb4
Nxd3+ 9.cxd3


Thanks

[mhalstern]

Miguel,

It is interesting that you are curious whether any wb engine works. I had tried joker114w (wb) which failed to run as well. The funny thing is that I can get joker working, pretending it is UCI, using a wb2uci adaptor! Unfortunatedly, this solution does not work with gaviota.
I'm looking forward to the new UCI Version. Post a link when it is ready. I'll let you know if this works with Aquarium.

The Aquarium GUI has a lot of useful features, but it is a very very weird buggy product. As an example, Aquarium has a feature to export games to a pgn file. If I run a 2 game engine match from the same opening position and output the games to a pgn file, it will often output the eco code to one of the games, but not the other. The behavior is not consistent. As a programmer, I can't understand this. Features either work, or they don't work. How could they work sometimes?!


Thanks

[Teemu Pudas]

Gaviota worked without the ini.txt.

Comment out resign_mode = on.

[alpha123]

Miguel,

It is interesting that you are curious whether any wb engine works. I had tried joker114w (wb) which failed to run as well. The funny thing is that I can get joker working, pretending it is UCI, using a wb2uci adaptor! Unfortunatedly, this solution does not work with gaviota.
I'm looking forward to the new UCI Version. Post a link when it is ready. I'll let you know if this works with Aquarium.

The Aquarium GUI has a lot of useful features, but it is a very very weird buggy product. As an example, Aquarium has a feature to export games to a pgn file. If I run a 2 game engine match from the same opening position and output the games to a pgn file, it will often output the eco code to one of the games, but not the other. The behavior is not consistent. As a programmer, I can't understand this. Features either work, or they don't work. How could they work sometimes?!


Thanks

That tends to happen with Windows also.

I think I will actually buy Aquarium when Rybka 4 comes out. Thanks for the warning though.

Peter