Rodent 0.17 for New Year!

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

Moderators: hgm, Harvey Williamson, bob

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
PK
Posts: 755
Joined: Mon Jan 15, 2007 10:23 am
Location: Warsza
Contact:

Rodent 0.17 for New Year!

Post by PK » Mon Dec 31, 2012 11:02 am

Hi all!

It has been a busy year for me, including bad throat inflammation last January, starting a relationship, ending a relationship, moving to various parts of Warsaw 4 times, changing internet provider, making a streak of 26 tournament games without a loss and beating a couple of candidate masters in the process, suffering under the burden of translation of XVIIth century Latin letters etc. It didn't leave me much time for programming, so I decided to make an interim release of Rodent 0.17 as it is today. I don't expect a strength increase. New Rodent is available at:

http://www.pkoziol.cal24.pl/rodent/rodent.htm

changes include
- null reduction depth bugfix (spotted again by invaluable Kestutis Gasaitis)
- some tinkering with aspiration window
- candidate passers eval
- some bugfixes in checkmate eval and material eval

As usual, I have to wait for good people to supply me with 64-bit and Linux compilers.

Happy New Year!

PK
Posts: 755
Joined: Mon Jan 15, 2007 10:23 am
Location: Warsza
Contact:

Re: Rodent 0.17 for New Year!

Post by PK » Mon Dec 31, 2012 11:06 am

A comparison with the very first version of Rodent: same depth, same speed, no chance:

[pgn]
[Event "RodentYear"]
[Site "EURO-KOMPUTER"]
[Date "2012.12.31"]
[Round "1"]
[White "Rodent"]
[Black "Rodent_old"]
[Result "1-0"]
[BlackElo "2200"]
[ECO "A80"]
[Opening "Dutch"]
[Time "11:00:18"]
[Variation "2.Nc3 Nf6 3.Bg5 e6"]
[WhiteElo "2200"]
[TimeControl "300+1"]
[Termination "adjudication"]
[PlyCount "77"]
[WhiteType "human"]
[BlackType "human"]

1. d4 f5 2. Bg5 Nf6 {(Ng8f6 Bg5xf6 e7xf6 e2e3 Nb8c6 Bf1d3 d7d5 Qd1f3 g7g6
Nb1c3 Bc8e6 Ng1e2 Bf8e7) -0.26/13 8} 3. Nc3 d5 {(d7d5 Bg5xf6 e7xf6 e2e3
g7g6 Bf1d3 c7c6 Ng1f3 Bf8d6 OO OO Qd1d2 Bc8e6 Qd2e2) -0.27/14 8} 4. f3 Nc6
{(Nb8c6 e2e3 e7e6 Bf1d3 h7h6 Bg5h4 Bf8d6 Ng1e2 Nc6b4 Nc3b5 Nb4xd3+ Qd1xd3
Bd6e7 Bh4g3 c7c5) +0.01/14 8} 5. Qd2 e6 {(e7e6 e2e3 h7h6 Bg5h4 g7g5 Bh4g3
Bf8d6 Bf1b5 Bd6xg3+ h2xg3 Bc8d7 Ng1e2 Qd8e7 g3g4 f5xg4) +0.10/14 6} 6.
O-O-O Bd7 {(Bc8d7 Bg5xf6 g7xf6 e2e3 Bf8d6 Ng1h3 a7a6 Bf1d3 f5f4 Nh3xf4
Bd6xf4 e3xf4 Nc6xd4) +0.11/13 7} 7. e3 {(e2e3 h7h6 Bg5f4 g7g5 Bf4g3 Bf8d6
Ng1h3 Qd8e7 Bg3xd6 c7xd6 Nc3b5 OOO Bf1d3) +0.04/14 12} Be7 {(Bf8e7 Bf1d3 OO
a2a3 h7h6 Bg5h4 g7g5 Bh4g3 Be7d6 Bg3xd6 c7xd6 h2h4 e6e5 h4xg5 h6xg5)
+0.01/14 7} 8. Nh3 {(Ng1h3 OO Bf1e2 a7a6 Bg5h4 Nf6e8 Bh4xe7 Nc6xe7 Nh3f4
Ne8d6 Kc1b1 b7b6 Nf4d3 Nd6c4) +0.09/14 6} O-O {(OO Bg5h4 h7h6 Bh4g3 g7g5
Kc1b1 a7a6 Rd1e1 Nf6h5 Bg3f2 Be7d6 Bf1d3 Nc6b4 Bd3e2 Nb4c6) +0.15/15 6} 9.
Be2 {(Bf1e2 a7a6 Kc1b1 b7b6 Rh1f1 h7h6 Bg5f4 g7g5 Bf4g3 Nf6h5 Bg3f2 Nh5f6
Be2d3) +0.06/13 6} h6 {(h7h6 Bg5h4 g7g5 Bh4g3 a7a6 a2a3 Qd8c8 Qd2d3 Nf6h5
Bg3f2 Be7d6 e3e4 g5g4 e4xd5 e6xd5 f3xg4 f5xg4) +0.25/15 6} 10. Bh4 {(Bg5h4
a7a6 Bh4g3 Nf6h5 Bg3e1 Be7b4 g2g4 f5xg4 f3xg4 Nh5f6 Nh3f4 Nf6e4 Qd2d3
Ne4d6) +0.22/14 10} g5 {(g7g5 Bh4g3 a7a6 Rd1e1 b7b6 Be2d3 Qd8c8 Kc1b1 Qc8b7
a2a3 Nf6h5 Bg3f2 Be7d6 e3e4) +0.16/14 7} 11. Be1 {(Bh4e1 Nc6a5 Nh3f2 Na5c4
Be2xc4 d5xc4 h2h4 g5xh4 e3e4 Nf6h5 Nf2h3 f5xe4 f3xe4) +0.29/13 8} a6 {(a7a6
Nh3f2 b7b5 a2a3 Nc6a5 h2h4 Na5c4 Be2xc4 b5xc4 h4xg5 h6xg5 e3e4) +0.12/12 7}
12. Nf2 {(Nh3f2 Be7d6 h2h4 Nf6h5 Nf2d3 f5f4 h4xg5 Qd8xg5 Be1h4 f4xe3 Bh4xg5
e3xd2+ Bg5xd2 Nh5g3 Rh1xh6 Nc6xd4) +0.46/13 7} b5 {(b7b5 a2a3 Be7d6 h2h4
Nf6h5 Nf2d3 f5f4 h4xg5 Qd8xg5 Be1f2 Nh5g3 e3xf4 Ng3xe2+ Qd2xe2 Qg5xg2
Rh1xh6 Bd6xf4+ Nd3xf4 Rf8xf4) -0.01/12 6} 13. h4 {(h2h4 b5b4 Nc3a4 Nf6h5
Nf2d3 f5f4 Na4c5 f4xe3 Qd2xe3 Nh5f4 Nd3xf4 g5xf4 Qe3d3 Be7xc5 d4xc5)
+0.57/14 7} gxh4 {(g5xh4 Nf2d3 Nf6h5 Rh1h3 b5b4 Nc3b1 Rf8f7 f3f4 Nh5f6
Be1xh4 Nf6e4 Bh4xe7 Qd8xe7) -0.01/13 5} 14. Nd3 {(Nf2d3 Nf6h5 Kc1b1 f5f4
Nd3xf4 Nh5xf4 e3xf4 Be7f6 Be1f2 Qd8e7 Bf2e3 Qe7g7 Rh1g1 Nc6b4) +0.60/14 7}
Nh5 {(Nf6h5 f3f4 Nh5f6 Be1xh4 b5b4 Nc3b1 Nf6e4 Qd2e1 Be7xh4 Rh1xh4 Qd8f6
Qe1h1 Ne4g3) -0.16/13 6} 15. f4 {(f3f4 Nh5f6 Be1xh4 b5b4 Nc3b1 Nf6e4 Qd2e1
Be7xh4 Rh1xh4 Qd8f6 Nb1d2 Qf6g6 Nd2xe4 f5xe4) +0.92/14 7} Be8 {(Bd7e8 Kc1b1
Nc6a5 Nd3c5 Be7xc5 d4xc5 b5b4 Nc3a4 Nh5g3 Be1xg3 h4xg3 Qd2xb4) -0.23/12 6}
16. Bf3 {(Be2f3 b5b4 Nc3e2 Nc6a5 b2b3 Na5b7 Nd3xb4 Nb7d6 c2c4 Nd6e4 Bf3xe4
f5xe4 Kc1b1 d5xc4 b3xc4) +0.96/14 7} Ng7 {(Nh5g7 Qd2e2 b5b4 Nc3b1 a6a5
Nd3c5 h6h5 Nb1d2 a5a4 Kc1b1 Be7xc5 d4xc5 Qd8e7) -0.30/13 6} 17. Nc5 {(Nd3c5
Be7xc5 d4xc5 Qd8e7 Be1xh4 Qe7xc5 Bh4f2 b5b4 Nc3e2 Rf8f6 Bf2h4 Rf6g6 Kc1b1
e6e5 f4xe5) +1.07/14 6} Bf7 {(Be8f7 Nc3e2 Be7f6 Be1f2 Qd8e7 Qd2e1 Bf7h5
Bf3xh5 Ng7xh5 Bf2xh4 Bf6xh4 Qe1xh4 Qe7xh4 Rh1xh4) -0.43/13 6} 18. Bf2
{(Be1f2 Be7xc5 d4xc5 Qd8e7 Rh1xh4 h6h5 Rh4h3 b5b4 Nc3a4 Kg8h7 c2c3 b4xc3
Na4xc3 Qe7xc5) +1.09/13 6} Qe8 {(Qd8e8 Nc5d3 Bf7h5 Bf2xh4 Bh5xf3 g2xf3 b5b4
Nc3e2 h6h5 Bh4xe7 Qe8xe7 Nd3e5 Nc6xe5 f4xe5) -0.39/13 6} 19. Bxh4 {(Bf2xh4
Be7xc5 d4xc5 Bf7h5 Kc1b1 b5b4 Nc3e2 Qe8f7 Rh1f1 Bh5xf3 g2xf3 d5d4 Ne2xd4
Nc6xd4 Qd2xd4) +1.12/15 9} Bxc5 {(Be7xc5 d4xc5 Bf7h5 Rd1g1 Kg8h8 Kc1b1
Qe8f7 Bh4f2 Kh8g8 Qd2e2 Bh5xf3 Qe2xf3 b5b4 Nc3e2) -0.45/14 5} 20. dxc5
{(d4xc5 Bf7h5 Kc1b1 b5b4 Nc3e2 Qe8f7 Rh1f1 Bh5xf3 g2xf3 d5d4 Ne2xd4 Nc6xd4
Qd2xd4) +1.12/14 9} Bh5 {(Bf7h5 Rd1g1 Kg8h8 Kc1b1 Qe8f7 Bh4f2 Qf7g6 Rh1h3
b5b4 Nc3e2 Bh5xf3 g2xf3 Qg6f6) -0.47/13 5} 21. Rdg1 {(Rd1g1 Kg8h8 Nc3e2
Qe8g6 Ne2g3 Bh5g4 Bh4g5 h6h5 Bf3d1 Rf8f7 Qd2d3 Bg4xd1 Rg1xd1 e6e5 Qd3xd5)
+1.24/14 8} Kh8 {(Kg8h8 Kc1b1 Qe8f7 Nc3e2 Bh5xf3 g2xf3 Ng7h5 Rg1g2 Rf8g8
Qd2c3+ Kh8h7 Rh1g1 Rg8xg2 Rg1xg2) -0.49/12 5} 22. Ne2 {(Nc3e2 Qe8g6 Kc1b1
Kh8g8 Ne2c3 Ra8e8 Qd2f2 Kg8h8 Nc3e2 Bh5g4 Bf3xg4 f5xg4 Rg1d1) +1.18/13 5}
Qf7 {(Qe8f7 Kc1b1 Bh5g4 Bh4g5 h6h5 Rg1f1 Kh8g8 b2b3 Bg4xf3 Rf1xf3 e6e5
f4xe5 Nc6xe5) -0.61/12 5} 23. Be1 {(Bh4e1 Qf7g6 Qd2d3 Kh8h7 Rh1h4 Bh5xf3
g2xf3 Qg6f6 Rh4g4 Rf8g8 Be1c3 d5d4 Ne2xd4 Nc6xd4 Bc3xd4) +1.41/13 7} Qg6
{(Qf7g6 Qd2d3 Bh5xf3 g2xf3 Qg6h7 Rh1h2 Rf8g8 Rg1h1 h6h5 Be1c3 d5d4 Ne2xd4
Nc6xd4 e3xd4) -1.03/13 5} 24. Qd3 {(Qd2d3 Kh8h7 Bf3xh5 Ng7xh5 g2g4 Nh5g7
Rh1h4 Ra8c8 Rg1h1 h6h5 g4xh5 Qg6h6 Be1c3) +1.57/13 4} Bxf3 {(Bh5xf3 g2xf3
Qg6h7 Be1c3 Rf8f7 Rg1g2 Nc6e7 Rh1g1 Ra8g8 Kc1b1 Ne7c6 b2b4 d5d4 Ne2xd4
Nc6xd4 Bc3xd4) -1.24/14 5} 25. gxf3 {(g2xf3 Qg6h7 Ne2d4 Nc6xd4 Qd3xd4 Rf8f7
Be1c3 a6a5 Rg1xg7 Rf7xg7 Qd4xg7+ Qh7xg7 Rh1xh6+ Kh8g8 Bc3xg7 Kg8xg7 Rh6xe6)
+1.86/15 5} Qh7 {(Qg6h7 Be1c3 Rf8f7 Rh1h2 d5d4 Ne2xd4 Ra8d8 Rh2g2 e6e5
f4xe5 Nc6xe5 Qd3e2 Rd8d5 b2b4 h6h5) -2.16/15 5} 26. Nd4 {(Ne2d4 Nc6xd4
Qd3xd4 Rf8f7 Be1c3 Ra8g8 Kc1d2 a6a5 a2a3 Rg8a8 Rg1xg7 Rf7xg7 Qd4xg7+ Qh7xg7
Rh1xh6+ Kh8g8 Bc3xg7 Kg8xg7 Rh6xe6) +1.91/14 8} Nxd4 {(Nc6xd4 Qd3xd4 Rf8f7
Be1c3 Ra8e8 Rg1g2 Kh8g8 Rh1g1 Re8e7 c5c6 h6h5 Bc3b4 e6e5 f4xe5) -1.40/14 5}
27. Qxd4 {(Qd3xd4 Rf8f7 Rg1g2 Ra8f8 Be1c3 a6a5 Rg2xg7 Rf7xg7 Qd4xg7+ Qh7xg7
Rh1xh6+ Kh8g8 Bc3xg7 Kg8xg7 Rh6xe6 Rf8g8) +2.22/14 5} Rf7 {(Rf8f7 Be1c3
Kh8g8 Rg1g2 Ra8e8 Rh1g1 Re8e7 c5c6 h6h5 Qd4a7 Qh7h8 Qa7xa6 d5d4 Qa6c8+
Re7e8) -1.76/14 5} 28. Rh2 {(Rh1h2 Kh8g8 Rh2g2 Ra8e8 Be1h4 Kg8f8 c5c6 Re8a8
Bh4e1 Ng7e8 Be1b4+ Ne8d6 Qd4c3 Kf8e7 Kc1d1) +2.19/14 4} Kg8 {(Kh8g8 Rh2g2
Ra8e8 Be1c3 Re8e7 c5c6 h6h5 Qd4a7 Kg8h8 Rg2g5 Qh7h6 Qa7xa6 h5h4 Qa6xb5 d5d4
Bc3xd4) -2.54/15 4} 29. Rhg2 {(Rh2g2 Ra8e8 Be1h4 Kg8f8 c5c6 Re8a8 Rg2g6
Ng7e8 Rg6xe6 Rf7g7 Rg1d1 Kf8g8 Qd4xd5 Kg8h8 e3e4) +2.54/15 4} Re8 {(Ra8e8
Be1c3 Re8e7 c5c6 Qh7h8 Qd4c5 Kg8h7 Bc3f6 Qh8e8 Bf6xe7 Qe8xe7 Qc5a7 h6h5
Qa7xa6 Kh7g8) -3.41/15 4} 30. Bh4 {(Be1h4 Kg8f8 c5c6 Re8b8 Bh4e1 Ng7e8
Be1b4+ Ne8d6 Qd4e5 Kf8e7 Rg2g6 Rb8f8 Bb4xd6+ c7xd6 Qe5xe6+ Ke7d8 Rg6xh6
Rf7e7 Rh6xh7 Re7xe6) +3.95/15 4} Kf8 {(Kg8f8 c5c6 Re8c8 Bh4e1 Ng7e8 Be1b4+
Ne8d6 Qd4e5 Rc8d8 Qe5xe6 h6h5 Qe6xd5 a6a5 Bb4xa5 Kf8e8) -2.89/14 4} 31. c6
{(c5c6 Re8c8 Qd4c5+ Kf8g8 Rg2g6 Kg8h8 Bh4f6 Rc8g8 Rg6xg7 Rf7xg7 Rg1xg7
Rg8xg7 Qc5f8+ Qh7g8 Bf6xg7+ Kh8h7 Bg7xh6 Qg8xf8 Bh6xf8) +4.65/15 4} Rc8
{(Re8c8 Bh4e1 Rc8e8 Qd4c5+ Rf7e7 Rg2xg7 Qh7xg7 Rg1xg7 Kf8xg7 Qc5a7 h6h5
Be1a5 h5h4 Ba5xc7 h4h3) -4.21/14 3} 32. Qc5+ {(Qd4c5+ Kf8g8 Rg2g6 Kg8h8
Bh4f6 Rc8g8 Rg6xg7 Rg8xg7 Rg1xg7 Rf7xg7 Qc5f8+ Qh7g8 Bf6xg7+ Kh8h7 Qf8xg8+
Kh7xg8 Bg7e5 d5d4 Be5xd4) +4.72/15 7} Kg8 {(Kf8g8 Rg2g6 Rc8f8 Bh4e7 Rf8e8
Be7f6 Rf7xf6 Rg6xf6 d5d4 Rf6g6 d4xe3 Rg6xg7+ Qh7xg7 Rg1xg7+ Kg8xg7 Qc5xe3
Kg7f7 Kc1d2) -4.91/15 4} 33. Rg6 {(Rg2g6 Kg8h8 Bh4f6 Rc8g8 Rg6xg7 Rg8xg7
Rg1xg7 Rf7xg7 Qc5f8+ Qh7g8 Bf6xg7+ Kh8h7 Qf8xg8+ Kh7xg8 Bg7e5 h6h5 Be5xc7
h5h4 Kc1d2) +5.40/16 5} Kh8 {(Kg8h8 Bh4f6 Rc8g8 Rg6xg7 Rf7xg7 Rg1xg7 Rg8xg7
Qc5f8+ Qh7g8 Bf6xg7+ Kh8h7 Qf8xg8+ Kh7xg8 Bg7e5 h6h5 Be5xc7 h5h4 Bc7e5 h4h3
c6c7) -5.47/17 4} 34. Bf6 {(Bh4f6 Rc8g8 Rg6xg7 Rg8xg7 Rg1xg7 Rf7xg7 Qc5f8+
Qh7g8 Bf6xg7+ Kh8h7 Qf8xg8+ Kh7xg8 Bg7e5 h6h5 Be5xc7 Kg8f8 Bc7e5 Kf8e8 c6c7
Ke8d7 Kc1d2) +5.97/18 5} Rg8 {(Rc8g8 Rg6xg7 Rf7xg7 Rg1xg7 Rg8xg7 Qc5f8+
Qh7g8 Bf6xg7+ Kh8h7 Qf8xg8+ Kh7xg8 Bg7e5 h6h5 Be5xc7 d5d4 e3xd4 h5h4 Kc1d2
h4h3) -5.54/16 4} 35. Rxg7 {(Rg6xg7 Rf7xg7 Rg1xg7 Rg8xg7 Qc5f8+ Qh7g8
Bf6xg7+ Kh8h7 Qf8xg8+ Kh7xg8 Bg7e5 Kg8f7 Be5xc7 Kf7e8 Bc7e5 h6h5 c6c7 Ke8d7
Kc1d2 h5h4 Kd2d3) +5.81/18 6} Rfxg7 {(Rf7xg7 Rg1xg7 Rg8xg7 Qc5f8+ Qh7g8
Bf6xg7+ Kh8h7 Qf8xg8+ Kh7xg8 Bg7e5 h6h5 Be5xc7 Kg8f7 Bc7e5 Kf7e7 c6c7 Ke7d7
Kc1d2 h5h4) -5.75/16 4} 36. Rxg7 {(Rg1xg7 Rg8xg7 Qc5f8+ Qh7g8 Bf6xg7+ Kh8h7
Qf8xg8+ Kh7xg8 Bg7xh6 Kg8f7 Bh6g5 Kf7e8 b2b4 Ke8f7 Bg5d8 Kf7g8 Bd8xc7 Kg8f7
Bc7e5 Kf7e7) +6.84/17 4} Rxg7 {(Rg8xg7 Qc5f8+ Qh7g8 Bf6xg7+ Kh8h7 Qf8xg8+
Kh7xg8 Bg7e5 Kg8f7 Be5xc7 Kf7e8 Bc7e5 h6h5 c6c7 Ke8d7 Kc1d2 h5h4 Kd2d3)
-5.76/15 4} 37. Qf8+ {(Qc5f8+ Qh7g8 Bf6xg7+ Kh8h7 Qf8xg8+ Kh7xg8 Bg7xh6
Kg8f7 Bh6g5 a6a5 Bg5d8 a5a4 Bd8xc7 Kf7e8 Bc7d6 Ke8d8 Kc1d2 b5b4 Kd2d3 a4a3
c6c7+ Kd8d7) +7.39/18 3} Qg8 {(Qh7g8 Bf6xg7+ Kh8h7 Qf8xg8+ Kh7xg8 Bg7e5
Kg8f7 Be5xc7) -5.49/4} 38. Bxg7+ {(Bf6xg7+ Kh8h7 Qf8xg8+ Kh7xg8 Bg7xh6
Kg8f7 Bh6g5 a6a5 Bg5d8 a5a4 Bd8xc7 Kf7e7 Bc7e5 Ke7e8 c6c7 Ke8d7 b2b3 a4a3
Kc1d2 b5b4) +7.45/18 5} Kh7 {(Kh8h7 Qf8xg8+ Kh7xg8 Bg7e5 Kg8f7 Be5xc7)
-5.49/4} 39. Qxg8+ {(Qf8xg8+ Kh7xg8 Bg7xh6 Kg8f7 Bh6g5 Kf7g8 Bg5d8 Kg8f7
Bd8xc7 Kf7e8 Bc7e5 a6a5 c6c7 Ke8d7 Kc1d2 Kd7c8 Kd2d3 Kc8d7) +7.48/17 3 1-0
User Adjudication} 1-0
[/pgn]

ZirconiumX
Posts: 1327
Joined: Sun Jul 17, 2011 9:14 am

Re: Rodent 0.17 for New Year!

Post by ZirconiumX » Mon Dec 31, 2012 11:27 am

Don't worry, Pawel, as soon as I can, I will make you some optimised compiles with ICC.

Matthew:out
Some believe in the almighty dollar.

I believe in the almighty printf statement.

User avatar
fern
Posts: 8745
Joined: Sun Feb 26, 2006 3:07 pm

Re: Rodent 0.17 for New Year!

Post by fern » Mon Dec 31, 2012 2:27 pm

Pavel, if you ever meet a polish lady named Lucyna falkiewicz-Wille, a scholar in language and so on, give her my salutes....

fern

PK
Posts: 755
Joined: Mon Jan 15, 2007 10:23 am
Location: Warsza
Contact:

Re: Rodent 0.17 for New Year!

Post by PK » Mon Dec 31, 2012 4:23 pm

bad luck, I have forgotten to change one line in search after an unsuccesful test -> just made a silent update. :oops:

User avatar
lucasart
Posts: 2957
Joined: Mon May 31, 2010 11:29 am
Contact:

Re: Rodent 0.17 for New Year!

Post by lucasart » Tue Jan 01, 2013 12:12 pm

PK wrote:Hi all!

It has been a busy year for me, including bad throat inflammation last January, starting a relationship, ending a relationship, moving to various parts of Warsaw 4 times, changing internet provider, making a streak of 26 tournament games without a loss and beating a couple of candidate masters in the process, suffering under the burden of translation of XVIIth century Latin letters etc. It didn't leave me much time for programming, so I decided to make an interim release of Rodent 0.17 as it is today. I don't expect a strength increase. New Rodent is available at:

http://www.pkoziol.cal24.pl/rodent/rodent.htm

changes include
- null reduction depth bugfix (spotted again by invaluable Kestutis Gasaitis)
- some tinkering with aspiration window
- candidate passers eval
- some bugfixes in checkmate eval and material eval
Thank you for your nice engine. Open source at its best: Pablo Vasquez made Sungorus, you improved it significantly into Rodent, maybe someone else will fork off Rodent, who knows ?

I remember Rodent was one tough sparring partner for DoubleCheck, when I started. In fact, at the beginning, even Sungorus was out of reach for DoubleCheck.
PK wrote: As usual, I have to wait for good people to supply me with 64-bit and Linux compilers.
I would gladly help, but you REALLY need to write portable code, otherwise it will be very hard to compile it. And even if I do, it would be a waste of time, as all the modifications I will have done to make it portable will be lost.
PK wrote:Happy New Year!
Happy new year to you too!
Theory and practice sometimes clash. And when that happens, theory loses. Every single time.

PK
Posts: 755
Joined: Mon Jan 15, 2007 10:23 am
Location: Warsza
Contact:

Re: Rodent 0.17 for New Year!

Post by PK » Tue Jan 01, 2013 12:52 pm

Lucas, perhaps it would be in accord with open source spirit, if You make portability changes, I release the source as 0.18 and will try my best not to spoil it afterwards?

I think that the best argument for open source (and for early release for that matter) is the help I obtained so far. In the most unusual form it came from a checker programmer, Kestutis Gasaitis, who spotted several bugs by just reading the code. Trying to do the same, I found two or three more.

As for strength, I'm afraid I begin banging my head against the wall. For sure some Elo can be gained in the eval, but the main problem is the search depth. Even though I concentrate on search modifications, it hardly increases.

BTW, would it be sensible to create a version of Rodent designed especially for forking - slightly weaker, but without non-standard stuff (for example with normal LMR) and without things like weakening algorithm?

ZirconiumX
Posts: 1327
Joined: Sun Jul 17, 2011 9:14 am

Re: Rodent 0.17 for New Year!

Post by ZirconiumX » Tue Jan 01, 2013 2:15 pm

Pawel, I would check out Ed Schröder's programmer's corner, especially the selective search section. Quiescence Search is irritating in the fact that there are only two main pruning techniques used there - that site gives a few more. That might help with your search depth...

Matthew:out
Some believe in the almighty dollar.

I believe in the almighty printf statement.

User avatar
Jim Ablett
Posts: 1327
Joined: Fri Jul 14, 2006 5:56 am
Location: London, England
Contact:

Re: Rodent 0.17 for New Year!

Post by Jim Ablett » Tue Jan 01, 2013 2:41 pm

PK wrote:Lucas, perhaps it would be in accord with open source spirit, if You make portability changes, I release the source as 0.18 and will try my best not to spoil it afterwards?

I think that the best argument for open source (and for early release for that matter) is the help I obtained so far. In the most unusual form it came from a checker programmer, Kestutis Gasaitis, who spotted several bugs by just reading the code. Trying to do the same, I found two or three more.

As for strength, I'm afraid I begin banging my head against the wall. For sure some Elo can be gained in the eval, but the main problem is the search depth. Even though I concentrate on search modifications, it hardly increases.

BTW, would it be sensible to create a version of Rodent designed especially for forking - slightly weaker, but without non-standard stuff (for example with normal LMR) and without things like weakening algorithm?

Code is already portable. Just add a few lines in 'rodent.h' such as >

Code: Select all

#include <stdint.h>

typedef uint64_t U64;
typedef uint32_t U32;
Jim.

User avatar
Jim Ablett
Posts: 1327
Joined: Fri Jul 14, 2006 5:56 am
Location: London, England
Contact:

Re: Rodent 0.17 for New Year!

Post by Jim Ablett » Tue Jan 01, 2013 7:24 pm

Here is the src code with makefile ported to GCC. Compiles with Mingw in Windows & Gcc in Linux & should in Osx as well.

https://dl.dropbox.com/u/5047625/rodent ... %20src.zip

I did an android compile that showed up a bug in Rodent. When run in 'Chess for Android' gui, the command 'go movetime' (time per move) doesn't function properly. The engine just searches forever and doesn't make a move.
It does this in Windows with Arena as well.

Jim.

Post Reply