A Gravity & Pressure revamp is needed

  • coryman
    30th Mar 2018 Member 0 Permalink

    @Ben_Ger (View Post)

     I get 30fps with a 1x1 grid, when the simulation is empty. Randomly using air and wind, putting a bunch of plut and neut, etc. brought it down to 20. My laptop isn't the best, but it's able to run games like Space Engineers without too much difficulty, so I'm sure other people could get anything from 10 to 45ish fps, depending on their computers.

    Also, for the record, making a version with a 1x1 grid is dead easy, as long as you know how to compile. You just find the value of cell size in one of the scripts and change 4 to 1, if memory serves

  • _Theo
    31st Mar 2018 Member 0 Permalink

    This would be pretty nice. Altough... even excluding the fps and toggleability and coding difficulties, 1x1 air would move very slowly

  • LeoTindall
    16th Apr 2018 Member 0 Permalink

    @jacob1 (View Post)

     I seriously doubt that it would slow the game. The value would be a single machine word (dword on 32 bits, qword on 64 bits) and would almost definitely remain in cache at all times, if not in a register (especially on x86_64, where we have the luxury of many general purpouse registers).

    Would you at least look at a PR if I were to submit one?

  • jacob2
    16th Apr 2018 Member 0 Permalink
    @LeoTindall (View Post)
    Changing CELL at runtime is still more work than just making it a variable instead of a define. It is possible it wouldn't have any effect on the generated code, but I feel like it would at least change a little (for example, it could probably do a bit shift to divide by cell right now)

    I still don't want cell to be variable. It should always be 4, except in mods which decide to change it. Feel free to code it, but unless it works out far better than I think it will probably be rejected.
    Edited 2 times by jacob2. Last: 16th Apr 2018