#77 Stockfish 17 running incorrectly in Chess 960

[Michael Brown]

Whereas Stockfish 17 runs well on 960 files in ChessX 1.6.2 and Lc0 31.1 runs well in ScidvsPC 960 (3533), Stockfish 17 runs incorrectly in ScidvsPC 960 (3533).
SF 17 starts OK, but after the next move in the game is selected, it doesn't show the new lines it is analysing. However the analysis box details are being updated (nodes, NPS, time, etc) and the log shows that SF 17 has been analysing the next move/moves.
When I press "stop" to halt the enigine, nothing happens.
There are no problems running SF 17 in ScidvsPC commit 3532.

[Michael Brown]

I should have said - running in Windows 11.

[Steven]

Please attach engine logs when reporting engine related issues.

I'm guessing your problem is now resolved, but had nothing to do with commit 3533.
Cheers

---- tcl/tools/uci.tcl  (revision 3532)
+--- tcl/tools/uci.tcl  (revision 3535)
 +++ tcl/tools/uci.tcl  (working copy)
 @@ -21,7 +21,7 @@

@@ -2433,7 +2433,7 @@
 +    if {[string match *UCI_Chess960* $analysis(uciOptions$n)]} {
 +      if {[lsearch $options UCI_Chess960] == -1} {
 +        ::splash::add "Enabling UCI_Chess960 for [lindex $engineData 0] (by default)."
-+        ::sendToEngine $n "setoption name UCI_Chess960 value 1"
++        ::sendToEngine $n "setoption name UCI_Chess960 value true"
 +      }
 +    } else {
 +      ::splash::add "Warning, UCI engine [lindex $engineData 0] does not support option UCI_Chess960"

[Michael Brown]

Yes, apologies about engine log. Although it is really above my pay grade, I have managed to use Tortoise to apply the revised chess960-new.patch in r3537 and compile it. Stockfish is now running normally in the 960 application.