#3744
"I believe mate in 500 (or more) is indeed possible, but only if one side made mistake earlier"
++ Yes, that is right. Here is an example: win in 400 moves.
We can prove black made a mistake earlier.
Black has 2 dark square bishops, so one must be an underpromoted pawn. The only reason to underpromote to a bishop instead of a queen is to avoid stalemate, i.e. to avoid a draw. As the position is lost for black, black made a mistake to underpromote to a bishop. He should have promoted to a queen and then the win in 400 is not there.
If we're going to disregard the rules that only exist to force every game to end in a reasonable timeframe, I'd love to see the collective world's computing power get stuck simulating a scenario in which both players move a knight out and back ad infinitum. As in, 1.Nf3 Nf6 2.Ng1 Ng8....
This is not some original Star Trek episode where the computer ends up exploding because you asked it "what is love?". It's incredibly easy to identify infinite repetitions, mark them as draws, and then proceed onwards.
I didn't say anything about exploding. Hanging up and requiring to be rebooted would be quite sufficient to satisfy my need for silly entertainment.