Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Crafty 11.23 in test (including comparison to commercial programs)

1 view
Skip to first unread message

Jouni Uski

unread,
Sep 17, 1997, 3:00:00 AM9/17/97
to

I run my own EPD-test suite with Crafty 11.23 (later versions don't
have EPD support). Test is tactically oriented, but still gives good
impression about strength of program. Here's Crafty's and others
results (PC is 486/150 = P90), bigger result is better:

Hiarcs 6, Mchess 6 138/192
Hiarcs 5 136
Rebel 8, Wchess 1.04 129
Rebel Decade 104
Comet A.63 103
CRAFTY 11.23 98

Crafty has very much problems in last test part, which contains
"trap" positions.

Jouni Uski

Mike Byrne

unread,
Sep 17, 1997, 3:00:00 AM9/17/97
to

On Wed, 17 Sep 1997 08:11:49 +0300, Jouni Uski <em...@address.com>
wrote:

>I run my own EPD-test suite with Crafty 11.23 (later versions don't
>have EPD support).

I don't understand this comment -- later versions do have epd support.
remove the "nospam." and substitute "com" for "spam" in the domain..
the spamming is out of control and even just the "nospam." is no
longer working

Robert Hyatt

unread,
Sep 17, 1997, 3:00:00 AM9/17/97
to

Mike Byrne (nospam....@voicenet.spam) wrote:

: On Wed, 17 Sep 1997 08:11:49 +0300, Jouni Uski <em...@address.com>
: wrote:

What he means is that since 12.0, the call to the EPD code has been
commented out. When I eliminated the buffered I/O in 12.0, the very
first thing someone found was that EPD seemd to "go ape" because of
the new I/O... It is possible to read things out of sequence since I
am now bypassing the normal I/O library buffering. When this came up,
I simply commented out the EPD call in option.c until I had time to
figure out what was going on...


Jouni Uski

unread,
Sep 19, 1997, 3:00:00 AM9/19/97
to

Jouni Uski wrote:
>
> I run my own EPD-test suite with Crafty 11.23 (later versions don't

> have EPD support). Test is tactically oriented, but still gives good
> impression about strength of program. Here's Crafty's and others
> results (PC is 486/150 = P90), bigger result is better:
>
> Hiarcs 6, Mchess 6 138/192
> Hiarcs 5 136
> Rebel 8, Wchess 1.04 129
> Rebel Decade 104
> Comet A.63 103
> CRAFTY 11.23 98
>
> Crafty has very much problems in last test part, which contains
> "trap" positions.
>
> Jouni Uski

Actually the EPD support is not so important: I can use instead of it
TEST and LOAD commands! And now I have tested Crafty 12.6 - it is much
faster than 11.23 and it got 126 points in my test suite!! Very good.
Also it solves WAC 230 in unbeliavable 19 minutes in my 486 PC...

Jouni Uski

Kai Luebke

unread,
Sep 19, 1997, 3:00:00 AM9/19/97
to

Jouni Uski (em...@address.com) wrote:

: Actually the EPD support is not so important: I can use instead of it


: TEST and LOAD commands! And now I have tested Crafty 12.6 - it is much
: faster than 11.23 and it got 126 points in my test suite!! Very good.
: Also it solves WAC 230 in unbeliavable 19 minutes in my 486 PC...

As I've said, IMO 12.6 is the strongest Crafty yet, 12.7 is weaker.
I'll probably wait for 13.x for to test next.

---
Shep

0 new messages