Today I learned that in chess it was once (theoretically) possible (not forbidden) to castle vertically. There was this chess journalist, Tim Krabbé, who composed a chess problem in 1972 in which the white king stood on its original square, e1, while a white pawn promoted to a rook on e8. Krabbé carefully reread the FIDE rules and discovered that they said the king and rook must not have moved before, but nowhere did they say that the rook had to be on the same rank as the king. The new rook on e8 had never moved? It hadn’t. Neither had the king. So, by the letter of the rules, the king could jump to e3, and the rook to e2. Vertical castling, formally flawless.
The problem spread throughout the chess world, everyone had a chuckle, and then in 1973 FIDE quietly amended the rules. Now they clearly state that castling can only be done with a rook on the same rank.
