Bug in Crafty 14.2 ??

Michael Diosi

unread,

Dec 10, 1997, 3:00:00 AM12/10/97

to

--------------971C6599A1F84B5AB282CB95
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Hello !!

I just started crafty (from the program epd2diag v1.2c-which i
downloaded from Rebel Site) and tested it with the mate in 30 position
posted a few days (1 week) ago to this group. After arriving depth 9
crafty starts counting & showing the evaluation, for same move every
second or so. This did not happend with the previous crafty versions. Is
this a bug or is this happening just to me.

The tested postion was : [Image]

Thanks in advance.
Michael.

--------------971C6599A1F84B5AB282CB95
Content-Type: multipart/related; boundary="------------CE4B643E120F533F66583784"

--------------CE4B643E120F533F66583784
Content-Type: text/html; charset=us-ascii
Content-Transfer-Encoding: 7bit

<HTML>
Hello !!

I just started crafty (from the program epd2diag v1.2c-which i downloaded 
from Rebel Site) and tested it with the mate in 30 position posted a few
days (1 week) ago to this group. After arriving depth 9
 crafty starts counting & showing the evaluation, for same move
every second or so. This did not happend with the previous crafty versions.
Is this a bug or is this happening just to me.

The tested postion was : <IMG SRC="cid:part1.348EB...@athene.informatik.uni-bonn.de" HEIGHT=304 WIDTH=304>

Thanks in advance.
 Michael.</HTML>

--------------CE4B643E120F533F66583784
Content-Type: image/jpeg
Content-ID: <part1.348EB...@athene.informatik.uni-bonn.de>
Content-Transfer-Encoding: base64
Content-Disposition: inline; filename="C:\WINDOWS\TEMP\nsmailKS.jpeg"

<encoded_portion_removed>
yf8Ajv0/VHU0UUVxn0gUUUUAf//Z
--------------CE4B643E120F533F66583784--

--------------971C6599A1F84B5AB282CB95--

Michael Diosi

unread,

Dec 10, 1997, 3:00:00 AM12/10/97

to

--------------01F79437D28EF9A304617582

Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Michael Diosi schrieb:

>
>
> Hello !!
>
> I just started crafty (from the program epd2diag v1.2c-which i
> downloaded from Rebel Site) and tested it with the mate in 30
> position posted a few days (1 week) ago to this group. After arriving
> depth 9
> crafty starts counting & showing the evaluation, for same move every
> second or so. This did not happend with the previous crafty versions.
> Is this a bug or is this happening just to me.
>
> The tested postion was : [Image]
>
> Thanks in advance.
> Michael.

Sorry this is again me!

This seems not to happen if i use crafty 14.2 standalone !
I did not have the time to check this before, ...
Michael

--------------01F79437D28EF9A304617582
Content-Type: multipart/related; boundary="------------076D541F7D312CEE80DF6886"

--------------076D541F7D312CEE80DF6886

Content-Type: text/html; charset=us-ascii
Content-Transfer-Encoding: 7bit

<HTML>
Michael Diosi schrieb:
<BLOCKQUOTE TYPE=CITE>

Hello !!

I just started crafty (from the program epd2diag v1.2c-which i downloaded 
from Rebel Site) and tested it with the mate in 30 position posted a few
days (1 week) ago to this group. After arriving depth 9
 crafty starts counting & showing the evaluation, for same move
every second or so. This did not happend with the previous crafty versions.
Is this a bug or is this happening just to me.

The tested postion was : <IMG SRC="cid:part1.348EE...@athene.informatik.uni-bonn.de" HEIGHT=304 WIDTH=304>

Thanks in advance.
 Michael.</BLOCKQUOTE>
  Sorry this is again me!

This seems not to happen if i use crafty 14.2 standalone !
 I did not have the time to check this before, ...
 Michael</HTML>

--------------076D541F7D312CEE80DF6886
Content-Type: image/jpeg
Content-ID: <part1.348EE...@athene.informatik.uni-bonn.de>

Content-Transfer-Encoding: base64
Content-Disposition: inline; filename="C:\WINDOWS\TEMP\nsmailKS.jpeg"

<encoded_portion_removed>
yf8Ajv0/VHU0UUVxn0gUUUUAf//Z
--------------076D541F7D312CEE80DF6886--

--------------01F79437D28EF9A304617582--

Kai Luebke

unread,

Dec 11, 1997, 3:00:00 AM12/11/97

to

Michael Diosi (di...@athene.informatik.uni-bonn.de) wrote:
:
: --------------971C6599A1F84B5AB282CB95
: Content-Type: text/plain; charset=us-ascii
: Content-Transfer-Encoding: 7bit
:
: Hello !!
:

<Insert standard reminder about posting binaries to non-binaries groups
here> !!!

: I just started crafty (from the program epd2diag v1.2c-which i

: downloaded from Rebel Site) and tested it with the mate in 30 position
: posted a few days (1 week) ago to this group. After arriving depth 9
: crafty starts counting & showing the evaluation, for same move every
: second or so. This did not happend with the previous crafty versions. Is
: this a bug or is this happening just to me.

I've noticed something similar under WinBoard. I came up with a position
where Crafty 14.2 was dumping the same move & eval like every 1/50 of a
second or so, with a "fail low" eval, like this:

3 0.50 -- Nf1
0.52 -- Nf1
0.54 -- Nf1

and so on.

If someone would like to check this, here's the PGN file up to move 19.
At move 20, Crafty (who had 4 book moves in this situation) started
tossing out the weird stuff above (book random 2).
After playing 20.Qf1 with 14.1, 14.2 continued the game with no problems.
(I use WinBoard 3.6.2 and everything has been stable so far with all
Crafties 12.0+).

[Event "Computer chess game"]
[Site "SHEP'S SERVER"]
[Date "1997.12.10"]
[Round "2"]
[White "Crafty 14.02"]
[Black "Rebel 8.0"]
[Result "0-1"]
[TimeControl "3600"]

1. e4 e5 2. Nf3 Nc6 3. Bb5 a6 4. Ba4 Nf6 5. O-O Be7 6. Re1 b5 7. Bb3 O-O 8.
c3 d5 9. exd5 Nxd5 10. Nxe5 Nxe5 11. Rxe5 c6 12. d4 Bd6 13. Re1 Qh4 14. g3
Qh3 15. Be3 Bg4 16. Qd3 Rae8 17. Nd2 Re6 18. a4 Qh5 19. axb5 axb5

---
Shep

Robert Hyatt

unread,

Dec 12, 1997, 3:00:00 AM12/12/97

to

Kai Luebke (klu...@MI.Uni-Koeln.DE) wrote:

: and so on.

: ---
: Shep

I can't reproduce this, because this entire game is in my book here,
and I don't know where your version dropped out. If you can email me
the log.nnn file, I'll see if I can reproduce it and see what's up.
I don't (haven't) see this on the chess servers, so it is obviously something
subtle...

--
Robert Hyatt Computer and Information Sciences
hy...@cis.uab.edu University of Alabama at Birmingham
(205) 934-2213 115A Campbell Hall, UAB Station
(205) 934-5473 FAX Birmingham, AL 35294-1170

Kai Luebke

unread,

Dec 15, 1997, 3:00:00 AM12/15/97

to

Robert Hyatt (hy...@crafty.cis.uab.edu) wrote:
: Kai Luebke (klu...@MI.Uni-Koeln.DE) wrote:
:
: : [Event "Computer chess game"]

: : [Site "SHEP'S SERVER"]
: : [Date "1997.12.10"]
: : [Round "2"]
: : [White "Crafty 14.02"]
: : [Black "Rebel 8.0"]
: : [Result "0-1"]
: : [TimeControl "3600"]
:
: : 1. e4 e5 2. Nf3 Nc6 3. Bb5 a6 4. Ba4 Nf6 5. O-O Be7 6. Re1 b5 7. Bb3 O-O 8.
: : c3 d5 9. exd5 Nxd5 10. Nxe5 Nxe5 11. Rxe5 c6 12. d4 Bd6 13. Re1 Qh4 14. g3
: : Qh3 15. Be3 Bg4 16. Qd3 Rae8 17. Nd2 Re6 18. a4 Qh5 19. axb5 axb5
:

: I can't reproduce this, because this entire game is in my book here,

: and I don't know where your version dropped out. If you can email me
: the log.nnn file, I'll see if I can reproduce it and see what's up.

It's in my book as well, but with "book random 2", 14.2 sometimes runs a
quick evaluation of the available book moves (I thought only book random 0
was supposed to do this), and that's what happens here at move 20.
But 14.2 just doesn't make a move.
The log file shows exactly what I said above, thousands of lines with

0.54 -- Nf1
0.56 -- Nf1
and so on. (the log file was about 800K after just 20 seconds!).

: I don't (haven't) see this on the chess servers, so it is obviously something
: subtle...

Yes, since this is the only occasion yet where such a thing popped up.

---
Shep

Robert Hyatt

unread,

Dec 15, 1997, 3:00:00 AM12/15/97

to

Kai Luebke <klu...@MI.Uni-Koeln.DE> wrote:

: Robert Hyatt (hy...@crafty.cis.uab.edu) wrote:
: : Kai Luebke (klu...@MI.Uni-Koeln.DE) wrote:
: :
: : : [Event "Computer chess game"]
: : : [Site "SHEP'S SERVER"]
: : : [Date "1997.12.10"]
: : : [Round "2"]
: : : [White "Crafty 14.02"]
: : : [Black "Rebel 8.0"]
: : : [Result "0-1"]
: : : [TimeControl "3600"]
: :
: : : 1. e4 e5 2. Nf3 Nc6 3. Bb5 a6 4. Ba4 Nf6 5. O-O Be7 6. Re1 b5 7. Bb3 O-O 8.
: : : c3 d5 9. exd5 Nxd5 10. Nxe5 Nxe5 11. Rxe5 c6 12. d4 Bd6 13. Re1 Qh4 14. g3
: : : Qh3 15. Be3 Bg4 16. Qd3 Rae8 17. Nd2 Re6 18. a4 Qh5 19. axb5 axb5
: :
: : I can't reproduce this, because this entire game is in my book here,
: : and I don't know where your version dropped out. If you can email me
: : the log.nnn file, I'll see if I can reproduce it and see what's up.

: It's in my book as well, but with "book random 2", 14.2 sometimes runs a
: quick evaluation of the available book moves (I thought only book random 0
: was supposed to do this), and that's what happens here at move 20.
: But 14.2 just doesn't make a move.

Looking at the code, book random 0 should be the only way that this can
happen... ie there is no other reason that Crafty will search a book move.
please email me the log.nnn file and I will look at what is going on.

: The log file shows exactly what I said above, thousands of lines with

: 0.54 -- Nf1
: 0.56 -- Nf1
: and so on. (the log file was about 800K after just 20 seconds!).

that usually means that the board had been destroyed, generally when the
king has been captured. The value of a king will leave the score outside
the alpha/beta window, no matter how wide it is set...

: : I don't (haven't) see this on the chess servers, so it is obviously something
: : subtle...

: Yes, since this is the only occasion yet where such a thing popped up.

: ---
: Shep