Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Getting a 6-man tablebase solution.

Author: J. Wesley Cleveland

Date: 11:44:27 10/11/99

Go up one level in this thread


On October 11, 1999 at 09:08:46, Robert Hyatt wrote:

>On October 11, 1999 at 03:55:14, blass uri wrote:
>
>>On October 11, 1999 at 00:01:56, Robert Hyatt wrote:
>>
>>>On October 10, 1999 at 21:17:44, Ted Sutton wrote:
>>>
>>>>How can one either generate or download from a web site a particular 6-man
>>>>tablebase: either for the appropriate forces, e.g., B+P vs. R+P, or better yet,
>>>>for an particular position?
>>>
>>>
>>>No, except for a couple of rare exceptions.  The files are _much_ bigger than
>>>you realize.  IE B+P vs R+P will be well over 4 gigs.
>>
>>It depends on the original position.
>>If the pawns are blocked in the original position and you generate tablebases
>>only for future position than it will be clearly less than 4 gig.
>>If you have all the 5 piece tablebases than the future position that you do not
>>know the outcome of them are only future position when you change the place of 4
>>pieces(2 kings and rook and bishop).
>>
>>Uri
>
>
>we don't have any code to produce 'partial' files, nor do we have any code to
>probe it.  Nor do we have any engine that does a test to see if it is in a
>special-case ending.  I don't see any reasonable use for partial files, except
>for a specific game.  The files take too long to build for a specific game, and
>the size is still significant, since the very symmetric files we already have
>are 400mb+ in size.  Adding pawns, even blocked ones, removes one level of
>symmetry that can be used.

If you think about it, any 6 piece ending with specific blocked pawns would have
<16M positions uncompressed.



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.