3D Version of TPT?

  • ItsMyStupidNick123
    22nd Oct 2016 Banned 0 Permalink
    This post is hidden because the user is banned
  • TPT_PL
    22nd Oct 2016 Member 4 Permalink
    @ItsMyStupidNick123 (View Post)
    There already is, go to the settings, et voila! You got 3D.
    OH NO, or you were talking about 3D version with a 3D particle grid?

    Huh, it would be very hard to code and none of the devs would make this. You know, that kind of game would be very slow, and too much work for too small profit.

    TL;DR Impossible, go outside and stop playing TPT. ^^
  • DanielGalrito
    22nd Oct 2016 Member 1 Permalink

    There are some 3D versions, except they aren't TPT, have a much smaller space, and less elements with less properties.

  • Sandwichlizard
    24th Oct 2016 Member 0 Permalink

    honestly.  it would be awesome.  it will take a group of serious devs money and dedication and it would have to be called somelthing else and be built ground up, completely seperate from TPT so it can be monetized.  that is the only way that A 3D version of this game is going to happen.  the debth of creative realization offered by TPT is far greater than any current 3D contemporaries.  it will take years of development to get close to what we would all want to play with.

  • techyman305
    24th Oct 2016 Member 2 Permalink

    Also, this has been rejected COUNTLESS times over the years, and the answer is the same every time.

  • jombo23
    25th Oct 2016 Member 1 Permalink

    Sandwichlizard:

    honestly.  it would be awesome.  it will take a group of serious devs money and dedication and it would have to be called somelthing else and be built ground up, completely seperate from TPT so it can be monetized.  that is the only way that A 3D version of this game is going to happen.  the debth of creative realization offered by TPT is far greater than any current 3D contemporaries.  it will take years of development to get close to what we would all want to play with.

     

     

    no no no a million times no!

     

    it cannot work because of the shear number of particles. what is so hard for people to understand about that

  • Sandwichlizard
    25th Oct 2016 Member 2 Permalink

    hey jombo23.  stop being a know it all asshole!  I think it could work with a gound up build.  Also TPT runs on a single core.  It does not take advantage of modern computer architecture.  When you look at what is bing done by others I think that the tech is there but not the desire.

  • techyman305
    25th Oct 2016 Member 0 Permalink

    @jombo23 (View Post)

     *sheer number. Also, you kind of have been being a jerk in a lot of places. Just be nice, man.

    Edited once by techyman305. Last: 25th Oct 2016
  • TPT_PL
    26th Oct 2016 Member 0 Permalink
  • jombo23
    26th Oct 2016 Member 2 Permalink

    Sandwichlizard:

    hey jombo23.  stop being a know it all asshole!  I think it could work with a gound up build.  Also TPT runs on a single core.  It does not take advantage of modern computer architecture.  When you look at what is bing done by others I think that the tech is there but not the desire.

     

     

    No, it couldnt, and im not being an asshole. This has been suggested a thousand times with the same response each fricking time.

     

    Right now its roughly 230k particles. 

     

    Lets say you add a second layer with layer interaction. Ok, we can move the second layer to another core, and then use a third core to calculate the interaction between the layers, plus the performance degredation of everything having to be synced and scheduled.

     

    if you want full 3d, assuming tpt is 607x379. thats 230,053 particles. If you want a full cube of an area, thats 139,642,171 particles, over 600 times the original amount, 607 times to be precise. Now, instead of having 2d simple physics, now theres 3d physics, which would be a whole lot harder to calculate, and there would have to be REAL collision between all of those entities. 

     

    it takes DAYS UPON DAYS to render a 10 million particle water simulation, and you want to increase that 14 times over. Get real.

     

    The engine itself has very little to do with it. Of course full 3d couldnt work with this one. Layers could though.

     

    You have to realize that it is just impossible. It would also take a tremendous amount of memory. At the moment, tpt takes ~50mb. 

     

    Now, according to people, the screen of TPT is actually a simple map of the particle IDs.

     

    So lets assume you have 140 million particles. assuming you have 1 byte per particle, thats 140 megs, but thats not the case.

     

    Particles have 4 byte velocities, both X, Y, and now Z. Particles have 4 byte X, Y and Z positions, particles have 2 or 4 byte tmp, tmp2, temp, ptype, ctype, flags, dcolor, pavg0, pavg1, life . So thats, 16 4byte pieces of data per particle, so 64 bytes. 64 bytes * 140 million particles = ~9gb of ram just to handle the raw particle data at all. 

     

    Seeing as how ram is roughly 5-20GB/s depending on how good of a computer youre on, you could at most get 2 frames of data a second ram wise, or half a frame of data per second on the low end. Thats also not counting a lot of overhead in the game itself, such as calculating and storing collision data for interactions between 140 million particles and how each one affects another.

     

     

    All of the realtime fluid simulations in tech demos you see these days are roughly50-100k particles, and the graphics cards it runs on, which are capable of 5-10 teraflops, thats 5-10 trillion floating point operations a second, struggles to maintain a reasonable fps

     

    it

     

    is

     

    not

     

    possible

     

    with 

     

    current

     

    technology

     

     

    Id also like to add in the fact that it has nothing to do with time, difficulty, or money

    Edited 2 times by jombo23. Last: 26th Oct 2016
Locked by Lockheedmartin: Rejected Idea