Daydreamer 1.7

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

Moderators: hgm, Rebel, chrisw

User avatar
Werner
Posts: 2871
Joined: Wed Mar 08, 2006 10:09 pm
Location: Germany
Full name: Werner Schüle

Re: Daydreamer 1.7

Post by Werner »

Hi,
how to setup the paths correct for
J:\gtb4 and
J:\gtb5

in uci options?
Werner
Aaron Becker
Posts: 292
Joined: Tue Jul 07, 2009 4:56 am

Re: Daydreamer 1.7

Post by Aaron Becker »

Werner wrote:Hi,
how to setup the paths correct for
J:\gtb4 and
J:\gtb5

in uci options?
Hi Werner,

To add multiple paths, separate them by semicolons in the option string. So, for your example, you should set the tablebase path to:
J:\gtb4\;J:gtb5\

I'm not totally sure whether or not this will work correctly if each of the paths doesn't have a trailing backslash. Please let me know if this doesn't work for you.
User avatar
michiguel
Posts: 6401
Joined: Thu Mar 09, 2006 8:30 pm
Location: Chicago, Illinois, USA

Re: Daydreamer 1.7

Post by michiguel »

Aaron Becker wrote:
Werner wrote:Hi,
how to setup the paths correct for
J:\gtb4 and
J:\gtb5

in uci options?
Hi Werner,

To add multiple paths, separate them by semicolons in the option string. So, for your example, you should set the tablebase path to:
J:\gtb4\;J:gtb5\

I'm not totally sure whether or not this will work correctly if each of the paths doesn't have a trailing backslash. Please let me know if this doesn't work for you.
The Gaviota's TB APIs should take care of the slash/backslash when it open the files.

Miguel
User avatar
Werner
Posts: 2871
Joined: Wed Mar 08, 2006 10:09 pm
Location: Germany
Full name: Werner Schüle

Re: Daydreamer 1.7

Post by Werner »

Hi,
I think it works here under Shredder GUI and Arena 2.01. But I have a lot of troubles testing it. Sometimes after setup a position under Arena the engine hangs.
I did setup a won 5men postiton and Daydreamer announces a mate score. When I continue playing with the other color the mate disappears and I think DD is not able to mate with Bishop and Knight.

For testing I even copied all gtb into one folder.
I´ll try now a eng-eng match with a 5men position

[Event "40 Züge in 2 min"]
[Site "Engine Match"]
[Date "2010.03.30"]
[Round "1"]
[White "Daydreamer JA 1.7 x64"]
[Black "Naum 4 x64 1CPU"]
[Result "1/2-1/2"]
[SetUp "1"]
[FEN "4k3/4p3/8/8/8/8/4N3/4KB2 w - -"]

1. Kd2 {+M28/4 0s} e5 {+M29/3 0s} 2. Ke3 {+4.27/38 9s} Ke7
{+M28/3 0s} 3. Nc3 {+4.30/36 4s} e4 {+M27/3 0s (Kd6)}
4. Nxe4 {+4.30/29 2s} Ke6 {+M27/3 0s} 5. Nc3 {+4.30/29 1s}
Ke5 {+M28/3 0s (Ke7)} 6. Bd3 {+4.30/31 1s} Kd6 {+M27/3 0s}
7. Kd4 {+4.30/29 1s} Ke6 {+M26/3 0s} 8. Nd5 {+4.30/30 1s}
Kf7 {+M25/3 0s (Kd6)} 9. Ke5 {+4.36/31 3s} Kg7 {+M24/3 0s}
10. Kf5 {+4.58/37 1s} Kf7 {+M24/3 0s} 11. Bb5 {+4.58/40 3s}
Kg7 {+M24/3 0s} 12. Be8 {+4.56/33 1s} Kf8 {+M25/3 0s (Kh6)}
13. Bh5 {+4.57/29 1s} Kg7 {+M24/3 0s (Kg8)} 14. Nf4
{+4.56/34 2s} Kh6 {+M23/3 0s} 15. Be8 {+4.54/26 1s} Kg7
{+M24/3 0s} 16. Ne6+ {+4.53/30 2s} Kh7 {+M24/3 0s (Kh8)}
17. Kf6 {+4.54/30 1s} Kh6 {+M23/3 0s (Kg8)} 18. Nf4
{+4.50/28 1s} Kh7 {+M22/3 0s} 19. Bc6 {+4.49/29 2s} Kh6
{+M21/3 0s} 20. Bd5 {+4.48/30 1s} Kh7 {+M20/3 0s} 21. Nh5
{+4.56/31 1s} Kh6 {+M21/3 0s} 22. Ng7 {+4.47/34 2s} Kh7
{+M22/3 0s} 23. Bf7 {+4.47/28 1s} Kh6 {+M23/3 0s} 24. Nf5+
{+4.47/27 1s} Kh7 {+M22/3 0s} 25. Kg5 {+4.46/31 4s} Kh8
{+M23/3 0s} 26. Nd4 {+4.46/31 3s} Kg7 {+M25/3 0s} 27. Bg6
{+4.46/29 1s} Kg8 {+M24/3 0s} 28. Kf6 {+4.45/26 1s} Kh8
{+M23/3 0s (Kf8)} 29. Nb5 {+4.46/32 1s} Kg8 {+M22/3 0s}
30. Be4 {+4.44/31 4s} Kh8 {+M21/3 0s (Kf8)} 31. Nd4
{+4.45/28 1s} Kg8 {+M22/3 0s} 32. Bd3 {+4.45/30 2s} Kh8
{+M23/3 0s (Kf8)} 33. Bc4 {+4.43/25 1s} Kh7 {+M22/3 0s}
34. Bd5 {+4.43/24 1s} Kh8 {+M23/3 0s (Kh6)} 35. Bf3
{+4.43/29 2s} Kh7 {+M23/3 0s (Kg8)} 36. Bc6 {+4.45/25 1s}
Kg8 {+M23/3 0s (Kh6)} 37. Nf5 {+4.43/25 1s} Kh7 {+M22/3 0s
(Kh8)} 38. Bf3 {+4.43/31 1s} Kg8 {+M22/3 0s} 39. Ng7
{+4.43/26 2s} Kh7 {+M23/3 0s} 40. Ne8 {+4.50/27 5s} Kh8
{+M23/3 0s (Kg8)} 41. Nd6 {+4.33/26 12s} Kg8 {+M22/3 0s}
42. Bd5+ {+4.36/19 8s} Kh7 {+M21/3 0s} 43. Nf7 {+4.32/21
10s} Kg8 {+M20/3 0s} 44. Kg6 {+3.82/20 9s} Kf8 {+M19/3 0s}
45. Kg5 {0.00/19 5s} Ke7 {+M22/3 0s (Kg8)} 46. Bc4 {0.00/20
4s} Kd7 {+M22/3 0s (Ke8)} 47. Kf6 {0.00/20 2s} 1/2-1/2
Werner
Elorejano
Posts: 101
Joined: Sat Mar 20, 2010 3:31 am

Re: Daydreamer 1.7

Post by Elorejano »

Hi, again!
First of all: DD now loads right egbb. Works fine.
With ctg books, things are equal. DD plays blue moves, even green moves are available.
Im downloading Perfect15.ctg for to experiment with this.
If you send me an email adress, i can send to you polyglots logs or maybe you prefer winboard debug file?
Tell me
Regards
User avatar
Graham Banks
Posts: 41412
Joined: Sun Feb 26, 2006 10:52 am
Location: Auckland, NZ

Re: Daydreamer 1.7

Post by Graham Banks »

Aaron Becker wrote:Ok, I've created a preview release of Daydreamer 1.75.
So for testing purposes, I should hold off for now?

Cheers,
Graham.
gbanksnz at gmail.com
User avatar
Werner
Posts: 2871
Joined: Wed Mar 08, 2006 10:09 pm
Location: Germany
Full name: Werner Schüle

Re: Daydreamer 1.7

Post by Werner »

Hm,
I am not sure - gtbs are not working correct, see here:

[Event "40 Züge in 2 min"]
[Site "Engine Match"]
[Date "2010.03.30"]
[Round "1"]
[White "Daydreamer 1.75 rc1 w32"]
[Black "Naum 4 x64 1CPU"]
[Result "1/2-1/2"]
[SetUp "1"]
[FEN "4k3/4p3/8/8/8/8/4N3/4KB2 w - -"]

1. Nc3 {+M28/4 0s} Kd7 {+M29/3 0s} 2. Kd2 {+M27/4 0s} Kd6
{+M28/3 0s} 3. Bc4 {+M26/4 0s} Ke5 {+M27/3 0s} 4. Ke3
{+M25/4 0s} Kf5 {+M26/3 0s} 5. Ne4 {+M24/4 0s} Ke5 {+M25/3
0s} 6. Bf7 {+M23/4 0s} e6 {+M24/3 0s} 7. Bg6 {+M22/4 0s}
Kd5 {+M23/3 0s} 8. Bh5 {+M23/4 0s} Kc4 {+M24/3 0s} 9. Bg6
{+M22/4 0s} e5 {+M23/3 0s} 10. Bh5 {+M21/4 0s} Kb4 {+M22/3
0s} 11. Be2 {+M21/4 0s} Ka4 {+M22/3 0s} 12. Kd2 {+M20/4 0s}
Kb3 {+M21/3 0s} 13. Bd3 {+M20/4 0s} Kb4 {+M21/3 0s} 14. Be2
{+M20/4 0s} Kb3 {+M21/3 0s} 15. Kd3 {+M19/4 0s} Kb4 {+M20/3
0s} 16. Bd1 {+M18/4 0s} Kb5 {+M19/3 0s} 17. Bg4 {+M19/4 0s}
Kb4 {+M20/3 0s} 18. Be6 {+M20/4 0s} Kb5 {+M21/3 0s} 19. Kc3
{+M19/4 0s} Ka5 {+M20/3 0s} 20. Bg4 {+M18/4 0s} Kb5 {+M19/3
0s} 21. Bd1 {+M17/4 0s} Kb6 {+M18/3 0s} 22. Ba4 {+M17/4 0s}
Kc7 {+M18/3 0s} 23. Bd1 {+M17/4 0s} Kb6 {+M18/3 0s} 24. Kc4
{+M16/4 0s} Kc7 {+M17/3 0s} 25. Ba4 {+M16/4 0s} Kd8 {+M17/3
0s} 26. Bb5 {+M15/4 0s} Ke7 {+M16/3 0s} 27. Kd5 {+M14/4 0s}
Kf7 {+M15/3 0s} 28. Be2 {+M14/4 0s} Ke7 {+M15/3 0s} 29. Bg4
{+M13/4 0s} Kf7 {+M14/3 0s} 30. Kd6 {+M12/4 0s} Kg7 {+M13/3
0s} 31. Bh5 {+M11/4 0s} Kf8 {+M12/3 0s} 32. Ke6 {+M10/4 0s}
Kg7 {+M11/3 0s} 33. Be8 {+M9/4 0s} Kf8 {+M10/3 0s} 34. Bf7
{+M8/4 0s} Kg7 {+M9/3 0s} 35. Ke7 {+M7/4 0s} Kh6 {+M8/3 0s}
36. Be8 {+M8/4 0s} Kg7 {+M9/3 0s} 37. Bh5 {+M9/4 0s} Kh6
{+M10/3 0s} 38. Bf7 {+M9/4 0s} Kg7 {+M10/3 0s} 39. Be8
{+M8/4 0s} Kh6 {+M9/3 0s} 40. Kf6 {+M7/4 0s} Kh7 {+M8/3 0s}
41. Bf7 {+M6/4 0s} Kh8 {+M7/3 0s} 42. Bc4 {+M5/4 0s} Kh7
{+M6/3 0s} 43. Ng5+ {+M4/4 0s} Kh8 {+M5/3 0s} 44. Bd5
{+M4/4 0s} e4 {+M5/3 0s} 45. Be6 {+M4/4 0s} e3 {+M5/3 0s}
46. Nf7+ {+M5/4 0s} Kh7 {+M6/3 0s} 47. Bf5+ {+M21/4 0s} Kg8
{+M23/3 0s} 48. Bd3 {+M21/4 0s} e2 {+M22/3 0s} 49. Nh6+
{+M22/4 0s} Kf8 {+M23/3 0s} 50. Bxe2 {+4.52/20 6s} Ke8
{+M22/3 0s} 51. Ke6 {+4.40/19 7s} Kf8 {+M23/3 0s (Kd8)}
52. Bh5 {+4.54/18 8s} Kg7 {+M24/3 0s} 53. Nf5+ {+4.49/19
8s} Kg8 {+M24/3 0s (Kf8)} 54. Kf6 {+4.46/17 6s} Kf8 {+M23/3
0s (Kh7)} 55. Bg6 {+4.56/19 10s} Kg8 {+M22/3 0s} 56. Nd4
{+4.45/16 5s} Kh8 {+M23/3 0s (Kf8)} 57. Be8 {+4.48/16 7s}
Kh7 {+M24/3 0s (Kg8)} 58. Nf3 {+4.48/17 5s} Kh6 {+M23/3 0s}
59. Nh4 {+4.48/18 6s} Kh7 {+M22/3 0s} 60. Bf7 {+4.48/17 4s}
Kh6 {+M21/3 0s} 61. Nf5+ {+4.48/16 5s} Kh7 {+M22/3 0s}
62. Ng3 {+4.48/17 5s} Kh6 {+M23/3 0s (Kh8)} 63. Bd5
{+4.46/16 6s} Kh7 {+M22/3 0s} 64. Be4+ {+4.51/16 5s} Kg8
{+M23/3 0s (Kh6)} 65. Bc6 {+4.51/18 10s} Kh7 {+M23/3 0s
(Kf8)} 66. Bd5 {+4.49/15 4s} Kh6 {+M23/3 0s} 67. Bg2
{+4.49/17 13s} Kh7 {+M23/3 0s} 68. Ne4 {+4.49/19 19s} Kh8
{+M23/3 0s} 69. Ng5 {+4.46/15 5s} Kg8 {+M22/3 0s} 70. Bc6
{+4.48/16 6s} Kh8 {+M21/3 0s (Kf8)} 71. Be4 {+4.48/16 6s}
Kg8 {+M20/3 0s} 72. Ne6 {+4.44/15 7s} Kh8 {+M21/3 0s}
73. Bg6 {+4.46/17 9s} Kg8 {+M22/3 0s} 74. Bf7+ {+4.47/17
6s} Kh7 {+M22/3 0s (Kh8)} 75. Kg5 {+4.46/17 5s} Kh8 {+M23/3
0s} 76. Bg6 {+4.45/18 7s} Kg8 {+M22/3 0s} 77. Kf6 {+4.58/19
9s} Kh8 {+M21/3 0s} 1/2-1/2
Werner
Aaron Becker
Posts: 292
Joined: Tue Jul 07, 2009 4:56 am

Re: Daydreamer 1.7

Post by Aaron Becker »

Thanks for the information, Werner. I'll figure this out and get it fixed before the 1.75 release.

Graham, right now I think the only testing to be done is for correctness issues like Wener and Carlos have pointed out. Testing its strength while it still has these problems would be a waste of your time. Once I have that all fixed up it and make a proper release it will be time to seriously evaluate its strength.
User avatar
Graham Banks
Posts: 41412
Joined: Sun Feb 26, 2006 10:52 am
Location: Auckland, NZ

Re: Daydreamer 1.7

Post by Graham Banks »

Aaron Becker wrote:Thanks for the information, Werner. I'll figure this out and get it fixed before the 1.75 release.

Graham, right now I think the only testing to be done is for correctness issues like Wener and Carlos have pointed out. Testing its strength while it still has these problems would be a waste of your time. Once I have that all fixed up it and make a proper release it will be time to seriously evaluate its strength.
That's what I thought Aaron. Thanks for confirming. :)

Cheers,
Graham.
gbanksnz at gmail.com
Elorejano
Posts: 101
Joined: Sat Mar 20, 2010 3:31 am

Re: Daydreamer 1.7

Post by Elorejano »

Ive tried with Perfect 15.ctg and works really fine. Seems that DD only like Sedat´s boks :P
Perfect 15 have this list of moves
1. e4!! (green)
1. d4 (blue)
1. c4 (red)

others books

1. e4!! (green)
1. d4! (green)
1. c4 !? (blue)
1. Nf3 ?! (blue)
1. b3? (red)

Maybe this particular colours and nags combination give confusion to DD?
Theres various ctg books freely availables, could you try with another than Perfect?
Your latest version seems very strong, perhaps you can optimize without those vey ugly dlls, like pthreads.
Thanks in advance