Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Some hard test positions for computer programs

Author: Robert Hyatt

Date: 19:56:42 02/26/99

Go up one level in this thread


On February 26, 1999 at 16:00:36, Jeremiah Penery wrote:

>On February 26, 1999 at 07:41:57, blass uri wrote:
>
>>I am interested in results of programs in these positions to decide which
>>program to use in correspondence chess.
>>
>>I am interested to know the results of Hiarcs7 with selectivity 5,6 and 7.
>>I want to know if selectivity 6 or 7 is realy better for correspondence games.
>>
>>These positions were posted here in the past.
>>
>>position number 3:
>>7r/1k2P2P/4p3/1pp5/8/2pp4/1n3PP1/R4K1R w - - 0 1
>>
>>The right move is g4 but programs(with the exception of exchess) usually prefer
>>the losing move e8=Q
>>
>>The best reply to g4 is Nc4 (probably black is losing but Nc4 is the best
>>chance) and programs usually have problems in finding Nc4.
>
>This is what Crafty (modified) thinks in this position.  Before this it liked
>the move Rh5, and it never thought about e8=Q.
>
>                    10->   5:14  -0.12   1. g4=R
>                    11     6:55     ++   1. g4!!
>                    11->   9:06   0.27   1. g4=R
>
>I don't know where "=R" is coming from, but I occasionally get strange output
>like this...Any ideas Bob?
>
>This is from a PII-300.
>
>Jeremiah


what can happen is that as it scrolls thru the ply-1 move list, it displays
them one by one.  If move number 5 is something like a8=R, and then move number
6 is g4, the =R might not get overwritten on the screen with some compilers.  I
supposedly add enough spaces to cover that up, but somehow it is possible to
get that kind of stuff...  I'll look to see if there is a safer way to
overwrite.

If you do "disp nomoves" do you still see anything like that, since it won't
display the ply-1 moves as they are searched, then..



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.