Ah, I understand now. You mean a separate, colour-coded collision mask that can be sampled at the player’s feet, rather than sprite-to-sprite collision.
GameCanvas does not currently have a dedicated collision-mask layer. TileLayer can mark complete tile types as solid using SetSolid() and query them with IsSolid(), but that is not fine-grained enough for slopes or different surfaces within a tile.
It is technically possible to load a matching mask using Picture.FromAsset() and query it with Pixel(x, y), but the current pixel-query implementation is not intended for repeated physics checks every frame, so I wouldn’t recommend that as the finished solution.
A useful addition would be an efficient collision-mask API that loads the mask once and lets you query a world coordinate. For example ColourAtPoint(x, y) or IsSolidAt(x, y). Different colours could represent solid ground, ladders, water, hazards, one-way platforms, and so on. We could also add a platformer example demonstrating foot probes and walking smoothly up slopes.
Is that the workflow you have in mind? Would each colour represent a different terrain type, or do you only need solid versus empty?
