Page 58 of 59

Re: 7-men Syzygy attempt

Posted: Sun Dec 02, 2018 6:19 pm
by Sesse
jdart wrote:
Sat Dec 01, 2018 2:04 am
I have looked recently at merging the CFish version of tbprobe.c into my fork of Fathom, but there a lot of diffs, so that it is kind of a tall order. And I don't have 7-man bases to test it with.
Yes, that's what I noticed, too…
Pull requests are welcome, though.
Does this hold in general? :-) I'd love to get in the patch in https://github.com/basil00/Fathom/issues/22.

Re: 7-men Syzygy attempt

Posted: Sun Dec 02, 2018 8:08 pm
by jdart
Does this hold in general?
Sure, you can submit it. I don't use this feature (move_to_str) myself - so I was not aware there was a bug there.

--Jon

Re: 7-men Syzygy attempt

Posted: Sun Dec 02, 2018 8:42 pm
by Ozymandias
syzygy wrote:
Sun Dec 02, 2018 6:16 pm
Ozymandias wrote:
Sat Dec 01, 2018 2:32 pm
syzygy wrote:
Sat Dec 01, 2018 1:27 pm
WDL for N pieces requires WDL for (N-1) pieces and below.
Got it, thanks.
What is not needed for WDL are the tables that can be reached through promotions.

So if you are only interested in KRPPvKRP and only need WDL, then you must download just KRPPvKRP and all 6-piece WDL TBs (or at least those that can be reached through captures, but you're not going to gain much by being selective with 6-piece tables if you are downloading the much bigger 7-piece tables). You don't need all KRXYvKRZ if you are only interested in adjudicating KRPPvKRP positions or in analysing positions for which KRPPvKRP is important.

(But if you want to play out KRPPvKRP positions to mate reliably, you need all the tables that can be reached through promotion and the corresponding DTZ tables.)
The total size of 6 men WDL is less than 1% the size of 7 men WDL, so cherrypicking those tables is rather excessive. Nonetheless, the explanation helps better understanding the inner workings of syzygy.

Re: 7-men Syzygy attempt

Posted: Sun Mar 10, 2019 11:55 am
by Jesse Gersenson
Any progress getting 7-man support for Fathom or cutechess?

Re: 7-men Syzygy attempt

Posted: Sun Mar 10, 2019 2:06 pm
by Nordlandia
Jesse Gersenson: Latest commit 984dff8 on 14 Jan (cutechess github repository)

7-piece adjudication is a requested feature.

I recall someone mentioned that Chessbase plans to give Fritz GUI's or Chessbase 7-piece syzygy support but that is few months back in time.

Re: 7-men Syzygy attempt

Posted: Sun Mar 10, 2019 3:09 pm
by jdart
I am going to look again at grafting it into Fathom. When I get time.

--Jon

Re: 7-men Syzygy attempt

Posted: Tue Mar 19, 2019 3:17 am
by jdart
One of the issues re Fathom and the CFish 7-man code is that the latter assumes you have a full legal move generator available - of course CFish has that, but it is GPL and I'd like to keep Fathom more permissively licensed, plus not all engines have/need/want a Stockfish-style movegen. So that implies a pluggable movegen and a fallback implementation as was done for other areas of Fathom. That is doable but not so easy.

--Jon

Re: 7-men Syzygy attempt

Posted: Tue Mar 19, 2019 10:04 pm
by syzygy
jdart wrote:
Tue Mar 19, 2019 3:17 am
One of the issues re Fathom and the CFish 7-man code is that the latter assumes you have a full legal move generator available - of course CFish has that, but it is GPL and I'd like to keep Fathom more permissively licensed, plus not all engines have/need/want a Stockfish-style movegen. So that implies a pluggable movegen and a fallback implementation as was done for other areas of Fathom. That is doable but not so easy.
But there is no difference between 6 men and 7 men in this respect.

I guess the bigger problem is that the Cfish code was rewritten/cleaned up, so it is not trivial to see what really has changed.

Re: 7-men Syzygy attempt

Posted: Wed Mar 20, 2019 9:40 pm
by konsolas
With regards to supporting 7-man syzygy in engines:

Is CFish the definitive source for the 7-man probing code?
Has the original code on syzygy1/tb been updated to support 7-man?

Re: 7-men Syzygy attempt

Posted: Wed Mar 20, 2019 11:54 pm
by jdart
konsolas wrote:
Wed Mar 20, 2019 9:40 pm
With regards to supporting 7-man syzygy in engines:

Is CFish the definitive source for the 7-man probing code?
I think so.
Has the original code on syzygy1/tb been updated to support 7-man?
No. See https://github.com/syzygy1/tb/issues/31