This commercial variant has only two unorthodox pieces, three if you count the initial pawn triple push and the generalized e.p. rule.
Nevertheless it is a grateful test case for an interactive diagram.
Because of its board shape, mentioned e.p. rules, and castling.
The possibility to black out squares already existed in the interactive diagram, although it was not widely advertized: when you define a piece named 'hole' it will black out the square wherever you place it. No piece can move to or slide over such a blacked-out square, nor can such squares be grabbed and moved. The Design Wizard currently does not provide a means to create holes; you would have to add the required line by typing it in the HTML code, like hole::::e4,e5,d4,d5 to blackout the board center of an 8x8 board.
The main problem here was to get the unorthodox e.p. capture right. The XBetza notation uses 'e' as modality to indicate e.p. capture, and this can be used as alternative or in addition to 'c' or 'm'. Its meaning is to capture the last-moved piece on an empty square it passed over. The diagram uses the convention that this can happen on each square where an initial-only non-jumping move (i.e. containing 'in') could have been blocked, but originally the 'n' modifier was implemented only on D or A atoms, testing only one square. This was fine for FIDE Pawns, but in Omega Chess we must deal with an ifmnH as well. The diagram code for handling 'n' has now been beefed up to test two squares for blocking, and test up to two e.p. squares for strictly forward pushes.
Castling
As for the castling with a non-corner piece: the diagram uses an algorithm that always allows castling with a piece that moves like an orthodox Rook, even if it is not a corner piece. Official XBetza notation for this would be isjO2, the 'j' on an O (= castling) atom having the specialmeaning that you 'jump' a square when looking for the castling partner, starting from the edge (or at least the nearest blacked-out square). The diagram description doesn't even need that, but of course would fail if there is a non-corner Rook on the King rank that you don't want to participate in castling.
Omega Chess
This commercial variant has only two unorthodox pieces, three if you count the initial pawn triple push and the generalized e.p. rule. Nevertheless it is a grateful test case for an interactive diagram. Because of its board shape, mentioned e.p. rules, and castling.
The possibility to black out squares already existed in the interactive diagram, although it was not widely advertized: when you define a piece named 'hole' it will black out the square wherever you place it. No piece can move to or slide over such a blacked-out square, nor can such squares be grabbed and moved. The Design Wizard currently does not provide a means to create holes; you would have to add the required line by typing it in the HTML code, like hole::::e4,e5,d4,d5 to blackout the board center of an 8x8 board.
The main problem here was to get the unorthodox e.p. capture right. The XBetza notation uses 'e' as modality to indicate e.p. capture, and this can be used as alternative or in addition to 'c' or 'm'. Its meaning is to capture the last-moved piece on an empty square it passed over. The diagram uses the convention that this can happen on each square where an initial-only non-jumping move (i.e. containing 'in') could have been blocked, but originally the 'n' modifier was implemented only on D or A atoms, testing only one square. This was fine for FIDE Pawns, but in Omega Chess we must deal with an ifmnH as well. The diagram code for handling 'n' has now been beefed up to test two squares for blocking, and test up to two e.p. squares for strictly forward pushes.
Castling
As for the castling with a non-corner piece: the diagram uses an algorithm that always allows castling with a piece that moves like an orthodox Rook, even if it is not a corner piece. Official XBetza notation for this would be isjO2, the 'j' on an O (= castling) atom having the specialmeaning that you 'jump' a square when looking for the castling partner, starting from the edge (or at least the nearest blacked-out square). The diagram description doesn't even need that, but of course would fail if there is a non-corner Rook on the King rank that you don't want to participate in castling.