New feature: Fluid Lean using Cell Phone Sensor

[No Trickster]

On 2/24/2020 at 3:17 PM, No Trickster said:

It takes 2 or 3 seconds I think "to build up", and after that "warming up" period it works great and is more responsive.

For example, that is fixed, no more "wasted time", so no more "gratuitous latency".  That was just a mishandling of the socket.  That "warming up" was related to OS buffers, like I said in one post, but was caused by mishandling the socket buffer.  Performance-wise:  I activated the FPS counter,  I uncapped FPS , I tried to reach more than 60 FPS but I couldn't (maybe my driver in linux is not uncapped or I didn't set vsync off ... I really didn't attack the "60 FPS problem")  what I wanted to see if this code was affecting FPS in any way and I found it does not.  Maybe 2 or 3 FPS , but that's not a big performance penalty.  I would need to try that by slowing the game FPS (forcefully), simulating a slower hardware, and try it going on 30 FPS to see if it is more affected.  Anyways, any other users testing it would be something good.

Right now, I  really don't know if Windows will have its own performance issues when porting this.  Because I made use of some features of sockets that maybe are linux-only or maybe not.  I didn't get there yet.  I hope I don't crash with a wall when I decide to port it to windows just because of "sockets stuff".

Edited March 8, 2020 by No Trickster