Crafty Bishop Mobility: Computed vs. Actual

[GothicChessInventor]

I have noticed that the precomputed sliding mobility of the Bishop in Crafty's code accounts only for "vacancies", not whether or not they are actual safe havens to plant the Bishop.

How difficult would it be to precompute mobility that also took into account the primary "robber" of mobility, enemy pawn squares? It would add a necessary dimension to the lookup, namely the bitmap of enemy pawns, but this could prove useful.

A bishop unmoved on c1 might look like it is mobile if it could reach d2, e3, f4, and g5, but if there were pawns on f6, e5, d4, and c3, is this Bishop truly mobile?

Just a thought.

[bob]

I have noticed that the precomputed sliding mobility of the Bishop in Crafty's code accounts only for "vacancies", not whether or not they are actual safe havens to plant the Bishop.

How difficult would it be to precompute mobility that also took into account the primary "robber" of mobility, enemy pawn squares? It would add a necessary dimension to the lookup, namely the bitmap of enemy pawns, but this could prove useful.

A bishop unmoved on c1 might look like it is mobile if it could reach d2, e3, f4, and g5, but if there were pawns on f6, e5, d4, and c3, is this Bishop truly mobile?

Just a thought.

you are using very old code, although the current version does not do this either. It is very expensive for the gain...