Robert Nürnberg
Robert Nürnberg
### Describe the issue Stockfish playing as black hangs the queen in the position `r5k1/1p2qpb1/2p1bnp1/r3p2p/7Q/P1NPPB1P/1BP3P1/1R3R1K b - -` by suggesting as bestmove `f6e8` ### Expected behavior Stockfish should suggest any...
Might need a second pair of eyes, and some testing. Locally all seems fine.
This PR implements the changes discussed in https://github.com/tmaklin/bxzstr/pull/34. This compiles and works for my local project, but apart from that no extensive testing was done.
This PR offers a cleaner implementation of the new `istreambuf` constructor introduced in https://github.com/tmaklin/bxzstr/pull/32. We also add a new overloaded `ifstream` constructor, allowing the simpler syntax ``` bxz::ifstream input(file, bxz::zstd);...
This PR is a follow up to #6422. We now also abort DTZ probes at root if we run low on time. ``` > ./fastchess -engine name="patch" cmd=../stockfish/src/stockfish.patch -engine name="master"...
### Describe the issue In the recent TCEC game https://tcec-chess.com/#div=sf&game=65&season=28 Stockfish reported a mate in 5 for the position at 87b. The fen is `r5k1/5q2/1pP3P1/3Pp2P/1B2Pp2/5Q1K/8/8 b - - 0 87`....
As per the [PGN specification](https://www.saremba.de/chessgml/standards/pgn/pgn-complete.htm#c8.2), this should not happen. See also https://github.com/cutechess/cutechess/issues/172. I wrote this simple "one move engine" that can be used to test the correct implementation in fastchess:...
This PR replaces #905. Resuming/starting from a config file is now handled better, as well as initialization of the crc value.
This PR introduces additional checks for PVs reported by engines: * PVs for mate scores (without upperbound/lowerbound) are checked for completeness (PV leads to checkmate at the announced distance) *...