pietrko wrote:It is unphysical - should i explain?
In simple model collision with a wall and human body (at angle) is non-elastic (this is direct translaiton of polish word, the meaning is that the energy isn't conserverd in collision and some kinetic energy is wasted and transferred into heat and you don't bounce off).
So given that non-elastic collision your velocity:
- in direction normal to the surface of the wall would be 0 - so you woudld be stopped (but only in one direction)
- tangent direction: same as before but after that probably fading beacuse of friction with wall.
So before collision your velocity ( V_N - velocity compnent normal to the wall, V_T -tangent )
V_before = V_N + V_T
immidiately after
V_after = V_T, while later V_T fading due to friction.
In real life situation, since human is soft body and not a material point so there would be also some rotatation.... ehh well you can try it yourself
Actually depening on what you wear (how elastic colliding body is) bouncing is also possible in extreme case (perfectly elastic):
V_after = -V_N + V_T
there is no physical scenario during which:
(assuming: V_T != 0)
V_before = V_N + V_T
V_after = 0
Which is game behaviour.
But you devs are ok with movement so this is end of story.
Just don't say that it is pretty physical beacuse it is not - I didn't lie or made up this argument.
Of course I just gave simple model.
You shouldn't let a math-based degree get in the way of real life experience. Most of that stuff is calculated in a friction-free vacuum with point bodies and perfect inertia and other concessions to making the math easier. What you describe is also not 'physical'.
I prefer to rely on decades of experience caroming off walls and running into door frames which I've painstakingly gathered by never watching where the hell I'm going. When I run into something I rarely bounce off and I have never slid along the surface, rotating slightly.