lawyershost.blogg.se

1. #Endgame tablebase 7 pieces generator
2. #Endgame tablebase 7 pieces update
3. #Endgame tablebase 7 pieces code

Sample position: 1k6/3K1p2/8/8/8/8/P5P1/8 w -įixed bug with duplicate moves when position contained a black bishop.įixed problem where a illegal position could make some tablebases inaccessible. This would sometimes return incorrect results in positions where a double pawn moves was possible but did not lead to e.p. You may send your suggestions and feedback through our feedback pageįixed bug en-passant status bug.

Positions where castling is still possible are not in the Tablebases, therefore entering a position where castling is possible gives a result as if castling is not possible.

If the server has not been used for some time, the initial response may take a bit long. After that you will receive the new position with the new scores and moves.įor programmers the service is also accessible from within your own application. You may carry out one of the moves in the list by selecting the move of choice and pressing submit or by double clicking the move. Result is the distance to mate or 0 if the position is a draw. After a short time the textbox will show the result for the position and results for the possible moves. To use it just enter a FEN string in the box above and press the submit button or setup a position by dragging the pieces. All 3,4, and 5 piece endgames are available. This is an experimental Nalimov Tablebase Server. For a workaround see this article: /en/download/help/ java_blocked.xml

#Endgame tablebase 7 pieces update

This applet may not work if you have Java version 7 update 51, in that case a security error may be shown.

#Endgame tablebase 7 pieces generator

One problem is that the standard nalimov generator doesn't work with 7-men tablebases (I have no clue why not).Feb 2014. I will see if I can find it back.Īnd I read an estimate that the 7-men tablebases will be about 200Tb big. One (based on depth to conversion, rather then depth to mate) included an ending which took just over 500 (!) moves until a piece was finally taken. I came across a website with some people working on 7-men tablebases. And I am not sure if a 32-men tablebase would increase interest in chess a lot Since 6-men tablebases take 5-men tablebases take 7 Gb, and 6-men tablebases about 1 Tb, I assume 7-men would be in the order of magnitude of at least 100 Tb (perhaps even a lot more because of exponential growth). People are still working on 7 men tablebases, but not sure there's much progress. Generating most 6 piece tablebases is relatively fast, if I understood correctly, and they are available. Right now, time is the biggest issue in adding both DTC and 6-piece tables, I'm juggling a few other feature additions that will keep me away from endgames for a while. Most of the ones where people say, "I don't care if my move takes 10 moves longer to mate, it converts to something trivial much quicker" have pawns).

#Endgame tablebase 7 pieces code

I've noticed a new potential source of DTC (depth to conversion) tables which I might be able to use (someone else also provided some code for DTC, but in its current form it only produced pawnless endgames, and the reasons I'd like to have DTC is mainly for the endgames with pawns, as it allows me to be more lenient on some human like moves where the human move might take quite a few more moves to mate but is the fastest way to convert to a simpler endgame. I'd still really like to have 6 piece tablebases, even in the short time since this thread started the price per GB has fallen dramatically, so price is not the major barrier, but rather time in modifying some of the code to deal with extra pieces properly.