Software for solving chess moremovers

Discussion of anything and everything relating to chess playing software and machines.

Moderator: Ras

MOBMAT
Posts: 401
Joined: Sat Feb 04, 2017 11:57 pm
Location: USA

Re: Software for solving chess moremovers

Post by MOBMAT »

Does this engine run under UCI protocol?
i7-6700K @ 4.00Ghz 32Gb, Win 10 Home, EGTBs on PCI SSD
Benchmark: Stockfish15.1 NNUE x64 bmi2 (nps): 1277K
chessica
Posts: 981
Joined: Thu Aug 11, 2022 11:30 pm
Full name: Esmeralda Pinto

Re: Software for solving chess moremovers

Post by chessica »

Hello, I just installed the software and entered a small test configuration.

Fen: 7k/5pp1/8/8/8/8/1K6/R7 w - - 0 1
#18

1. I can't enlarge the window; it's too small for me.
2. It doesn't find a solution within 9 minutes.
msterkowiec
Posts: 15
Joined: Sat Apr 26, 2025 7:01 pm
Full name: Marcin Sterkowiec

Re: Software for solving chess moremovers

Post by msterkowiec »

@MOBMAT: no, no support to UCI protocol - the engine for now is supposed to communicate only with the predefined, built-in UI (WinAPI or wxWidgets) that also controls user licence (Free/Light/Complete).

@chessica: You must have missed the part marked "important" about downloading 5-men endgame tablebases (~4GB; see https://jgisland.pl/download/linux ) and, what's more, you must have been offline: the outcome of 5-piece positions is taken either from data in Precalc folder or downloaded from Syzygy (see Options - in many cases it may be worth increasing the level of lookup to Syzygy - it is 1 by default not to block engine's threads on download from internet). Then the mentioned 7k/5pp1/8/8/8/8/1K6/R7 w - - 0 1
#18 is solved immediately.

As far as the window size is concerned: I hope no part was truncated. If the window seems too small on some systems, at this moment the only possibility is to change the screen resolution (at least temporarily).
chessica
Posts: 981
Joined: Thu Aug 11, 2022 11:30 pm
Full name: Esmeralda Pinto

Re: Software for solving chess moremovers

Post by chessica »

Hello and good morning, where do I have to enter the path for the egtb ( syzygy ) I can't find it.
msterkowiec
Posts: 15
Joined: Sat Apr 26, 2025 7:01 pm
Full name: Marcin Sterkowiec

Re: Software for solving chess moremovers

Post by msterkowiec »

Hello, good morning. At this moment this path is hardcoded (https://tablebase.lichess.ovh) and not configurable. You can only configure how extensively the engine queries this service: line 7 in Options dlg - see the tooltip over this field:
The engine can automatically take advantage of 7-piece Syzygy tablebases (although it is not fully suitable for evaluation of moremovers, since this database contains only WDL/WinDrawLoss data) by quering internet database. This involves a slowdown, which can be considered significant, that's why this setting was introduced to reduce latencies.
The value of this settings specifies positions for how many first white moves should be queried in this Syzygy/Lichess database.
By default only the original position of a problem is queried in Syzygy/Lichess database (1 level). By doing this the engine is able to immediately skip the analysis of moves that lead to draw or lose.
Note also that every position downloaded from Syzygy/Lichess is stored in subfolder SavedTmpSyzygyData of program installation folder. However these local databases are stored per problem, so the same position may have to be looked up more than once in Syzygy/Lichess, if it appears during analysis of different starting positions.
BTW: you can find sample queries for example here: https://github.com/lichess-org/lila-tablebase, e.g.:
curl http://tablebase.lichess.ovh/standard?f ... _w_-_-_0_1
curl http://tablebase.lichess.ovh/standard/m ... _w_-_-_0_1
chessica
Posts: 981
Joined: Thu Aug 11, 2022 11:30 pm
Full name: Esmeralda Pinto

Re: Software for solving chess moremovers

Post by chessica »

A bit of cpu jumps out of my PC. I will therefore never accept these hard-coded tables. I made an exception once. Never again.
msterkowiec
Posts: 15
Joined: Sat Apr 26, 2025 7:01 pm
Full name: Marcin Sterkowiec

Re: Software for solving chess moremovers

Post by msterkowiec »

Well, actually it's on the contrary: endgame tablebases let save processor usage. Unless you mean these several bytes in RAM, which, if configurable, would also occupy... several bytes. My decision not to make this path configurable was due to the fact that there is still strong dependency on the data format, so that such flexibility would actually be illusory.
To sum up, I must admit I didn't get your point - please describe the issue you mean in a few more words - maybe then it will be possible to find some solution or workaround.
chessica
Posts: 981
Joined: Thu Aug 11, 2022 11:30 pm
Full name: Esmeralda Pinto

Re: Software for solving chess moremovers

Post by chessica »

Dann Corbit wrote: Wed Apr 30, 2025 2:02 am Bit of an unusual chess set you've got with the following:

[d]7k/8/6Q1/5ppp/4pppp/4pqnn/2p1pqnr/2Kbrrbn w - -
...

Code: Select all

Deep Blue:
 ...
 77/98	00:08,945	 37.146k	4.153k	+M24	Dg6-f6+ Kh8-h7 Df6-f8 Kh7-g6 Df8-h8 Kg6-f7 Dh8-h7+ Kf7-f6 Dh7-g8 Kf6-e7 Dg8-g7+ Ke7-e8 Dg7-f6 Ke8-d7 Df6-e5 Kd7-d8 De5-e6 Kd8-c7 De6-e8 Kc7-d6 De8-f7 Kd6-c6 Df7-e7 Kc6-b6 De7-d7 Kb6-c5 Dd7-c7+ Kc5-d4 Dc7-d6+ Kd4-c4 Dd6-c6+ Kc4-d3 Dc6-d7+ Kd3-c4 Dd7-d6 Kc4-b5 Dd6-c7 Kb5-a6 Dc7-b8 Ka6-a5 Db8-b7 Ka5-a4 Db7-b6 Ka4-a3 Db6-b5 Ka3-a2 Db5-b2+
 78/74	00:09,438	 38.952k	4.127k	+M25	Dg6-f6+ Kh8-h7 Df6-f8 Kh7-g6 Df8-h8 Kg6-f7 Dh8-h7+ Kf7-f6 Dh7-g8 Kf6-e7 Dg8-g7+ Ke7-e8 Dg7-f6 Ke8-d7 Df6-e5 Kd7-d8 De5-e6 Kd8-c7 De6-d5 Kc7-c8 Dd5-d6 Kc8-b7 Dd6-d8 Kb7-c6 Dd8-e7 Kc6-b6 De7-d7 Kb6-c5 Dd7-c7+ Kc5-d4 Dc7-d6+ Kd4-c4 Dd6-c6+ Kc4-d3 Dc6-d7+ Kd3-c4 Dd7-d6 Kc4-b5 Dd6-c7 Kb5-b4 Dc7-c6 Kb4-b3 Dc6-c5 Kb3-a4 Dc5-b6 Ka4-a3 Db6-b5 Ka3-a2 Db5-b2+
 79/50	00:09,630	 39.742k	4.127k	+M25	Dg6-f6+ Kh8-h7 Df6-f8 Kh7-g6 Df8-h8 Kg6-f7 Dh8-h7+ Kf7-f6 Dh7-g8 Kf6-e7 Dg8-g7+ Ke7-e8 Dg7-f6 Ke8-d7 Df6-e5 Kd7-d8 De5-e6 Kd8-c7 De6-d5 Kc7-c8 Dd5-d6 Kc8-b7 Dd6-d8 Kb7-c6 Dd8-e7 Kc6-b6 De7-d7 Kb6-c5 Dd7-c7+ Kc5-d4 Dc7-d6+ Kd4-c4 Dd6-c6+ Kc4-d3 Dc6-d7+ Kd3-c4 Dd7-d6 Kc4-b5 Dd6-c7 Kb5-b4 Dc7-c6 Kb4-b3 Dc6-c5 Kb3-a4 Dc5-b6 Ka4-a3 Db6-b5 Ka3-a2 Db5-b2+
User avatar
Ajedrecista
Posts: 2152
Joined: Wed Jul 13, 2011 9:04 pm
Location: Madrid, Spain.

Re: Software for solving chess moremovers.

Post by Ajedrecista »

Hello Dann:
Dann Corbit wrote: Wed Apr 30, 2025 2:02 amBit of an unusual chess set you've got with the following:

[d]7k/8/6Q1/5ppp/4pppp/4pqnn/2p1pqnr/2Kbrrbn w - -
[...]
Pretty much like other positions already posted, like the one of this thread and the 'Untouchable King' problem of 1969 (post 1 and post 2 for more context), this time with less available squares for the black king. This kind of problems is incredible.

Regards from Spain.

Ajedrecista.
msterkowiec
Posts: 15
Joined: Sat Apr 26, 2025 7:01 pm
Full name: Marcin Sterkowiec

Re: Software for solving chess moremovers

Post by msterkowiec »

Version 9.0 of J.G.Island is now ready for download at https://jgisland.pl. Among others, this version contains a new feature that may be important for many users: generating a file with solution details. This feature is configurable, although by default no more than 10 first moves are logged/displayed (and maximum file size is 1MB; of course in a typical case the size will be no more than one or two "screens", sometimes it is a single line - just the main variant)

@Ajedrecista
J.G.Island has precalculated solution data (an approach based on Nalimov endgame tablebases) for positions like the one you mentioned in your post. Position details tab appears then - and for the position 7k/8/6Q1/5ppp/4pppp/4pqnn/2p1pqnr/2Kbrrbn it displays:
Qg6-f6 #24
Qg6-h6 #25
Qg6-a6 #26
Qg6-b6 #26
Qg6-c6 #26
Qg6-d6 #26
Qg6-e6 #26
Qg6-e8 #26
and no mate or n.a. (effectively: no mate) for all other moves.