Author: Mike Byrne
Date: 12:02:01 07/17/04
Go up one level in this thread
On July 17, 2004 at 13:12:38, Drexel,Michael wrote: >On July 17, 2004 at 10:50:04, Robert Hyatt wrote: > >>On July 17, 2004 at 01:39:09, Graham Banks wrote: >> >>>That sounds horrible. No pondering. >>> >>> >>>Yes, I'm really hoping that one day I'll be able to own a dual processor machine >>>as I'd much prefer to use ponder on. Crafty definitely performs much better with >>>ponder on in comparison to other programs. >>>Still, beggars can't be choosers! >>>Graham. >> >> >>That wasn't my point. Mike or someone said the adapter sends a "new" command >>for each move. that will (a) kill any pondering; (b) clear the hash table >>before each move making it impossible to pass things along from one search to >>another; (c) slow the search by clearing the pawn hash, making it necessary to >>do lots of extra pawn evaluations; etc... > >Seems correct. >I just startet a game against Crafty 19.15 in the Fritz GUI. >The Wb2uci logfile has only 2 "new" commands at the start of the game but with >logfile enabled I got 10 logfiles in the Crafty directory. >For example: > >White(1): post >White(1): hard >pondering enabled. >White(1): computer >playing a computer! >White(1): level 0 4 0 >White(1): force >White(1): d4 >Black(1): d5 >White(2): c4 >Black(2): c6 >White(3): Nf3 >Black(3): Nf6 >White(4): Nc3 >Black(4): Be6 >White(5): cxd5 >Black(5): Bxd5 >White(6): e3 >Black(6): e6 >White(7): Bd3 >Black(7): Bxf3 >White(8): Qxf3 >Black(8): Be7 >White(9): O-O >Black(9): O-O >White(10): Bd2 >Black(10): Qa5 >White(11): Ne4 >Black(11): Qc7 >White(12): Nxf6+ >Black(12): Bxf6 >White(13): Rac1 >Black(13): Nd7 >White(14): time 27100 >White(14): otim 26198 >White(14): go > clearing hash tables > time limit 6.77 (47.39) > depth time score variation (1) > 6-> 0.10 0.86 14. Bb4 Rfd8 15. Qh5 g6 16. Qf3 Bg7 > 7 0.17 0.77 14. Bb4 Rfd8 15. Qh5 g6 16. Qf3 a5 > 17. Bd2 > 7 0.65 0.83 14. Qh3 g6 15. Be4 Be7 16. Qh6 Nf6 > 17. Bd3 > 7-> 0.74 0.83 14. Qh3 g6 15. Be4 Be7 16. Qh6 Nf6 > 17. Bd3 > 8 0.94 0.74 14. Qh3 g6 15. Qg4 Be7 16. e4 Nf6 17. > Qf4 Qxf4 18. Bxf4 > 8 1.37 0.81 14. Qe4 g6 15. Qf3 Qd6 16. Be4 Be5 > 17. h3 Bh2+ 18. Kh1 Be5 19. Kg1 > 8-> 2.09 0.81 14. Qe4 g6 15. Qf3 Qd6 16. Be4 Be5 > 17. h3 Bh2+ 18. Kh1 Be5 19. Kg1 > 9 4.41 0.75 14. Qe4 g6 15. a4 Be7 16. b4 Rad8 17. > b5 Ba3 18. Rc4 Nf6 > 9 5.05 0.77 14. Qh3 g6 15. f4 Rfd8 16. Qg4 Be7 > 17. Rfe1 Nf6 18. Qg5 > time=7.89 cpu=95% mat=0 n=8738682 fh=93% nps=1.11M > ext-> chk=200169 cap=19552 pp=5035 1rep=29979 mate=1857 > predicted=0 nodes=8738682 evals=1772393 50move=2 > endgame tablebase-> probes=0 hits=0 >White(14): Qh3 > time used: 7.89 > > +---+---+---+---+---+---+---+---+ > 8 |<R>| . | | . | |<R>|<K>| . | > +---+---+---+---+---+---+---+---+ > 7 |<P>|<P>|<Q>|<N>| . |<P>|<P>|<P>| > +---+---+---+---+---+---+---+---+ > 6 | | . |<P>| . |<P>|<B>| | . | > +---+---+---+---+---+---+---+---+ > 5 | . | | . | | . | | . | | > +---+---+---+---+---+---+---+---+ > 4 | | . | |-P-| | . | | . | > +---+---+---+---+---+---+---+---+ > 3 | . | | . |-B-|-P-| | . |-Q-| > +---+---+---+---+---+---+---+---+ > 2 |-P-|-P-| |-B-| |-P-|-P-|-P-| > +---+---+---+---+---+---+---+---+ > 1 | . | |-R-| | . |-R-|-K-| | > +---+---+---+---+---+---+---+---+ > a b c d e f g h > >Black(14): g6 [pondering] > time limit 6.57 (45.99) > depth time score variation (8) > 8 0.25 0.70 15. f4 Rfd8 16. b4 Rac8 17. Be4 Bg7 > 18. Qg4 Nb6 > 8 1.05 0.72 15. a4 Rfd8 16. a5 Be7 17. Rfd1 a6 > 18. Be4 Nf6 > 8-> 1.55 0.72 15. a4 Rfd8 16. a5 Be7 17. Rfd1 a6 > 18. Be4 Nf6 > 9 5.05 0.69 15. a4 Rad8 16. Rfe1 Rfe8 17. a5 Ne5 > 18. Be4 Be7 19. Qh6 >Black(14): time 26500 >time remaining: 4:25 (crafty). >Black(14): otim 25873 >time remaining: 4:18 (opponent). >Black(14): g6 >predicted move made. > 9 7.20 0.71 15. f4 Rfd8 16. a4 Be7 17. b4 Nb6 18. > a5 Nd5 19. b5 > time=7.20 cpu=92% mat=0 n=7516770 fh=92% nps=1.04M > ext-> chk=109651 cap=20583 pp=2234 1rep=27145 mate=20 > predicted=1 nodes=7516770 evals=1895210 50move=3 > endgame tablebase-> probes=0 hits=0 > time used: 5.38 >White(15): f4 > time used: 1.86 > > +---+---+---+---+---+---+---+---+ > 8 |<R>| . | | . | |<R>|<K>| . | > +---+---+---+---+---+---+---+---+ > 7 |<P>|<P>|<Q>|<N>| . |<P>| . |<P>| > +---+---+---+---+---+---+---+---+ > 6 | | . |<P>| . |<P>|<B>|<P>| . | > +---+---+---+---+---+---+---+---+ > 5 | . | | . | | . | | . | | > +---+---+---+---+---+---+---+---+ > 4 | | . | |-P-| |-P-| | . | > +---+---+---+---+---+---+---+---+ > 3 | . | | . |-B-|-P-| | . |-Q-| > +---+---+---+---+---+---+---+---+ > 2 |-P-|-P-| |-B-| | . |-P-|-P-| > +---+---+---+---+---+---+---+---+ > 1 | . | |-R-| | . |-R-|-K-| | > +---+---+---+---+---+---+---+---+ > a b c d e f g h > >Black(15): Rfd8 [pondering] > time limit 6.57 (45.99) > depth time score variation (8) > 8 0.32 0.68 16. a4 a5 17. b3 Bg7 18. Qg4 Nf6 19. > Qg5 Nd5 > 8 0.94 0.69 16. b4 Rac8 17. a4 Bg7 18. Qg4 Nb6 > 19. a5 Nd5 > 8-> 1.29 0.69 16. b4 Rac8 17. a4 Bg7 18. Qg4 Nb6 > 19. a5 Nd5 > 9 2.25 0.71 16. b4 a6 17. Qg4 Bg7 18. a4 Nf6 19. > Qh3 Nd5 <HT> > 9-> 4.08 0.71 16. b4 a6 17. Qg4 Bg7 18. a4 Nf6 19. > Qh3 Nd5 <HT> > 10 6.01 0.72 16. b4 a6 17. Rf3 e5 18. fxe5 Nxe5 > 19. Rxf6 Nxd3 20. Rcf1 Rd7 21. Qg4 > 10-> 12.33 0.72 16. b4 a6 17. Rf3 e5 18. fxe5 Nxe5 > 19. Rxf6 Nxd3 20. Rcf1 Rd7 21. Qg4 > 11 29.51 0.69 16. b4 Qd6 17. Be4 Bg7 18. a3 Rac8 > 19. a4 Nf6 20. Bd3 Nd5 21. Rc5 > 11-> 56.41 0.69 16. b4 Qd6 17. Be4 Bg7 18. a3 Rac8 > 19. a4 Nf6 20. Bd3 Nd5 21. Rc5 > 12 1:13 0.67 16. b4 Qd6 17. a4 Nb6 18. a5 Nd5 19. > Rc5 b6 20. Rc4 Rd7 21. a6 Rad8 >Black(15): quit >LearnBook() executed the adapter does send a new command with every move under Fritz GUI, I have captured this is the winboard2uci log and it is also evidenced by all the log files - they are created because Crafty is receiving a "new" command with every move ...this does NOT happen under Arena as a UCI engine using the adapter ...I think your post after this shows the wb2uci log and it does show a new command with every move
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.