Just 'guy', I was the only one involved in that change. It seems that in Sonic 2, there are some hardcoded areas like that. Unless you accounted for those and had found an exception.
MarkeyJester's Sonic 128 - REVAMPED Now available on GitHub
#18
Posted 01 September 2014 - 06:35 PM
Yeah. The entire point of that commit was to make the system more S2-like. So, no bugfixes or the sort. Those would be made into a guide here that could applied to both S1-28 and S2.
What you're looking at shouldn't be a bug. You see, ObjHitWallLeft is most likely for non-player use only, and since the layer system doesn't apply to non-player objects, the additional checks aren't needed. What that code there is doing is checking the left/right collision bits(?) of the primary (default, no loop) collision.
What you're looking at shouldn't be a bug. You see, ObjHitWallLeft is most likely for non-player use only, and since the layer system doesn't apply to non-player objects, the additional checks aren't needed. What that code there is doing is checking the left/right collision bits(?) of the primary (default, no loop) collision.
This post has been edited by Clownacy: 01 September 2014 - 06:35 PM
#19
Posted 01 September 2014 - 11:31 PM
Clownacy, on 01 September 2014 - 06:35 PM, said:
Yeah. The entire point of that commit was to make the system more S2-like. So, no bugfixes or the sort. Those would be made into a guide here that could applied to both S1-28 and S2.
What you're looking at shouldn't be a bug. You see, ObjHitWallLeft is most likely for non-player use only, and since the layer system doesn't apply to non-player objects, the additional checks aren't needed. What that code there is doing is checking the left/right collision bits(?) of the primary (default, no loop) collision.
What you're looking at shouldn't be a bug. You see, ObjHitWallLeft is most likely for non-player use only, and since the layer system doesn't apply to non-player objects, the additional checks aren't needed. What that code there is doing is checking the left/right collision bits(?) of the primary (default, no loop) collision.
Good to know. I will double check the commit difference to my source and make sure they match up then. Thanks for that!
#22
Posted 02 September 2014 - 07:05 AM
The collision data itself is split like that (top solid in one bit, left/right/bottom solid in the other bit), so it can't do otherwise. But the way the check is done also matters -- finding walls check horizontally in collision data, finding ceilings check vertically up, etc.

05