Prove Me Wrong
From last time I decided to see how to get only the wall normals to return from the algorithm. However I first was thinking about if I could prove that this method could work or not. I then got into the assumption it wouldn’t work. But in the end I couldn’t prove it. Even with a normal from the entity itself, the collision would still work correctly.
Not much later after that realization I found myself debugging the code again. I wanted to get a better overview of it again. At the end of the train ride I had found several things wrong with the system. Not a surprise really, but there are so many things that are incorrect. When you think back to the old version, it also worked with rotation. So it’s definitely doable. I have to take it in steps again and actually spend time on it. Since I haven’t spend enough time on this problem considering the time I had available during the past weeks.
The 3DS port however is going nicely. One thing left to fix is the shader compilation. If that is sorted then I should have some very basic rendering working. From there I can start trying to render quads and actual textured quads. Pretty excited about that and probably very much on schedule to get a working port ready. So nothing to complain on that side for a change.