vvs-

I tested only directional jumps and sliding. I can't imagine how to test standing jumps in isolation. There just seems to be no place depending on that in original levels.

In the original she seems to be slower on the ascending trajectory and faster on the descending.

If a framerate is significantly lower than 60 then animations start to break down. This is because the original engine used fixed framerate and OpenTomb need to simulate that. Using...

The problem with the current system is not that it operates at a variable rate, but that the simulation resolution can fall below the acceptable level. When it happens, the...

Actually, it just shouldn't skip simulation step at all. Yes, that would cause it to visually slow down but at least it will do the correct thing.

There already was an attempt to do it right in `Engine_Frame`: ``` if(time > 0.1) { time = 0.1f; } ``` We need to fix it.

That constant is obviously wrong, that's why I said that we need to fix it. It should be at least 60 Hz.

That's why we can't "skip" simulation steps - it's too complicated. On the other hand, slowing the simulation down allows the engine to catch up properly without complex workarounds.

I didn't look at physics in that context. But the animations won't work if simulated speed falls below 60 Hz. That's exactly the reason for #37.

Here's a good [article](http://gafferongames.com/game-physics/fix-your-timestep/) which explains all of this.