Implementing SAT 2.0
During the train ride I could create a first pass of the new way of collision checking.
First I tried a slight modification to my version.
This did fix the issue with false positives somewhat.
But I decided to make a fresh piece of code that would take everything into account from the forum post that I found.
The actual checking of the collision is way simpler and smaller now.
However it doesn’t work correctly yet.
But that is almost always the case with a first pass.
So hopefully before the weekend I have had enough time to iron out all the initial bugs.
Then we can see if that forum post and my way of thinking are on the same line.