Making a tile collision system NOT to the power of 2

[daddyfatflab]

Hello! I've slowly been chipping away at a procedural generating game that uses tilesets and sprites to the dimensions of 24 x 16. To my understanding, all tile collision tutorials on youtube use the binary approach to snap your player object out of a tile. (using the bbox &~) Naturally, this approach doesn't work, so im curious of theres a workaround. What are your thoughts?

 

[TheouAegis]

(x div w) * w

Or

x ‐ (x mod w)

 

[daddyfatflab]

(x div w) * w

Or

x ‐ (x mod w)

can you go a little more into detail on how to implement this please?

 

[Rob]

Hello! I've slowly been chipping away at a procedural generating game that uses tilesets and sprites to the dimensions of 24 x 16. To my understanding, all tile collision tutorials on youtube use the binary approach to snap your player object out of a tile. (using the bbox &~) Naturally, this approach doesn't work, so im curious of theres a workaround. What are your thoughts?

If the tutorials use one variable for their calculations, use 2 instead (eg 24 for width, 16 for height)

 

[TheouAegis]

I am pretty sure they don't ALL use bitwise math, because I rarely see bitwise math around the forums these days.

x&~15 is fundamentally the same as x-=(x mod 16), and (x div 16)*16, both of which can be used with non-power-of-two dimensions.