Single Comment

Okay, there is complacent directionality, east and west, in castling and in en passant. Pawn ''in passing'' captures if sitting two forward sideways from a two-slide opening of opposite number to square orthogonally adjacent. Also well-understood is that the take is at the capturee's pass-over square. Insofar as arrays lack 4-way symmetry, e.p. can be effective to the left only, with to the right expressly prohibited. Likewise, castling, either ''free'' or fixed, ought to be tested Kingside only, expressly prohibiting Queenside. We ignore the <1% of CVs permitting eccentric castling with other than the cornered Rook (Sergey Sirotkin's Full Double) or more than two Rooks(Fred Lange's MegaChess). Neither is the intention to exclude some (minority of) Pawns from e.p. altogether. Instead ALL Pawns' e.p. is one-way same-way. Pawns need e.p. against opening two-sliders crucially for narrow boards (2x2x16 Racing) and unsymmetrical boards(Ramayana, Ultra Slanted Escalator). A fortiori, applicability is natural for to-be-captured Pawns opening three squares on 10x10, subject to e.p. Queenside castling has always been odd man/odd woman out. When board has 12 ranks, or 14, 16, 20, who has not cheaply counted over to the obscure long-castling squares? It never happens otherwise, so just get rid of them. Allusive conclusive: in selected applications one-side-only castling -- and e.p. in suitable contexts -- would be more logical, as well as in keeping with some notional natural/artificial Xiangqi palace, transported to western OrthoChess, there being but one. Well-implemented, in principle it begins to right the imbalances of unsymmetrical board and incompletely-symmetrical array. Intermezzo: (En passant: sleepers keepers) = (Castling: jeepers creepers) = (Stalemate: winners keepers) = (Checkmate: losers weepers).