1. This forum is obsolete and read-only. Feel free to contact us at support.keenswh.com

Should mass production be made easier?

Discussion in 'General' started by skarlitz, Feb 20, 2017.

Thread Status:
This last post in this thread was made more than 31 days old.
  1. skarlitz

    skarlitz Trainee Engineer

    Messages:
    18
    Merge blocks are pesky workarounds. Now that we have things like welders, rotors, pistons, programmable blocks, and scripts like MArmOS, perhaps Keen could kick something our way to make the automated construction of things in survival a little easier?

    Some ideas I've had:

    • The ability to merge blocks without the use of a merge block, by using welders. The idea I had was a non physical block, called a weld surface. It takes no materials, it's just a 2D "block" there to denote that the surface in question is weldable to another weld surface. You bring two weld surfaces near each other, like a merge block, and use a welder to "weld" them together. They then just snap together like any merge block. This would also carry over into blueprints, meaning you could do something like mass produce hull plates, use an arm to position them, and a welder to weld them into place. To make this a little easier, one could have the weld surface remove the border around the block, allowing you to slot things into place (currently you can't do something like pushing a single block through a hull, because both the single block and the hole have a border that make them too big to fit). This is my biggest suggestion.
    • A third grid size, extra small. This would make tiny welder arms possible, as well as tiny rovers and a whole slew of other devices. This is a bit far fetched, but I feel the current two sizes are a little restricting.
    • Ropes. Hoist things up. Perhaps make a spool block to wrap the rope around that attaches to a rotor? Ropes have been on my list for a while since stabilizing grids has become a hassle, usually requiring a Tug ship or a mechanical arm.
    That's all I've got for now, I'll update with more. Feel free to add your own, I'm rather curious what others have in mind.
     
    Last edited: Feb 20, 2017
  2. Burillo

    Burillo Junior Engineer

    Messages:
    648
    that's a cool concept. third grid was already possible at some point, but the "weld surface" sounds awesome.
     
  3. Stormigedon

    Stormigedon Trainee Engineer

    Messages:
    72
    Yeah a weld surface would be great. I always thought it was a bit silly that you couldn't just weld things together and needed a magic other block(that is vastly oversized on the small grid but that's an argument for another day)
     
  4. skarlitz

    skarlitz Trainee Engineer

    Messages:
    18
    Maybe we can get Keen in on it. It's really just a merge block that starts as soon as you weld on it, takes no resources, and goes away after use.

    Also it makes the sides of the block it's connected to slightly smaller so the part can be slid into place.
     
  5. Ronin1973

    Ronin1973 Master Engineer

    Messages:
    4,964
    There's a mod that allows for welding two grids together without a merge block. The 3rd size or (micro) grid topic has been done to death. A block on a grid is a block on a grid, so micro grids being built to the size of your average small block or large block construction would be a problem.
    Also, micro grids would be severely over-weight. Small block light armor already consumes one steel plate. So you would need an entirely new set of micro-components or fractional components. Block mass is calculated on the components within it.

    The physics of ropes and movable grids would be just awful to implement and consume A LOT of CPU cycles. Also ropes would have to be calculated for their strength based on length every time a force was applied to them. More CPU.
     
  6. skarlitz

    skarlitz Trainee Engineer

    Messages:
    18
    Just pitching yet another use for it.

    Or just allow people to use portions of steel plate. We already have fractional amounts of processed and raw materials. Doesn't seem like too much trouble in my mind.

    Absolutely not. Source Engine has demonstrated working rope physics running on third rate hardware from as far back as 2004.

    Take a look at Garry's Mod if you want a good idea of what actually takes up a lot of CPU time. We also already have ropes in the game in the form of rotor chains (though they'd be inferior to a hardcoded variant). So don't tell me it can't be done feasibly, even in KSH's engine, because it already has by the community.

    I'm also somewhat sure physics is done on the GPU but someone can feel free to prove me wrong.

    It's already been done in the form of rotors. With safety lock, it doesn't even torque your ship anymore, behaving as if it were native to the grid. It's not that big of a problem, in my opinion, considering it already exists in some fashion.
     
    Last edited: Feb 20, 2017
  7. ViroMan

    ViroMan Senior Engineer

    Messages:
    1,123
    Wait... I thought rope physics was something that could be completely handed off to GPU to handle... at least for Physx.
     
  8. Ronin1973

    Ronin1973 Master Engineer

    Messages:
    4,964
    Components are more than likely stored as integers. I believe the maximum value of an INT to be 2147483647. Chances are you won't have 2 billion steel plates lying around in one stack. That means everywhere an inventory of components is referenced in the game, that INT has to be changed to a value that can carry a decimal point or some sort of trickery has to happen. It would be far easier to create micro-components, as ingots already are calculated in fractional values.

    All of SE's physics are calculated on the CPU. Nothing is calculated on the GPU. So slinging your rope around will need to be calculated as well. In gravity, you're dealing with a pendulum that is going to exert a force on the ship it is attached to.

    The next problem is the precision of welding or grinding micro-blocks. Grinding small blocks is already hard enough to do with any accuracy. Try hitting a spot that's a fifth as small. That's going over like hiring Michael J Fox as your dentist.
     
  9. skarlitz

    skarlitz Trainee Engineer

    Messages:
    18
    I still maintain that ropes would not be CPU intensive, on the grounds that it's been done before both internally and externally. Supposing it's done right, anyways.

    As for grinder accuracy, good point. Though, KSH could make it adjustably accurate to compensate.

    Finally, the steel plates. That's more or less just a code refactor that could be automated.

    Or micro components. I'm not really against it, just providing an opposing viewpoint.
     
  10. Malware

    Malware Master Engineer

    Messages:
    9,867
    Not adding any kind if opinion here, just a general statement: When talking about features that uses any kind of computational power, people tend to forget to factor in the big picture. No, this-or-that feature does not require too much power on its own, but it is yet another thing that requires cpu/gpu power in addition to what's already there... There's only so much power available, and it must be rationed out. Is this-or-that feature really worth it or could that power be funneled somewhere more important?

    That said; nah. I don't think simplified rope physics would be all that much of a problem, and grids already move. Sure, maybe not the full-flegded system but a good-enough-does-its-job version, probably. Depends on how it's implemented. If it's a system that allows players to place as many active ropes as they want everywhere, well... that's kinda different...
     
    Last edited: Feb 21, 2017
  11. AutoMcD

    AutoMcD Senior Engineer

    Messages:
    2,369
    Weld surface: yes
    Also, would really love the ability to grab a small block and position it to said weld surface.

    Third grid size: no thanks. I'm of the opposite opinion, that everything should be small grid, and the large-grid blocks are just bigger blocks for the same grid. Like large cargo, etc.
     
  12. Renegrade

    Renegrade Trainee Engineer

    Messages:
    40
    Ew no thanks. I don't have to worry about whether or not the refinery has enough stuff, it will tell me when it runs out. I don't really want to have to remember if a block needs 0.033895 steel plates or not.

    Better block diversity would help with construction of smaller rovers and such. Most functional blocks are only available in a limited range of scales, usually one or two sizes at most, and are usually vastly oversize (ex. a hydrogen tank is only available in 5x5x5 for small blocks). If they added more blocks, especially in the smaller range, it would be easier to construct smaller rovers and probes and such.

    Blocks could be compacted even further by combining some together. For example, an RCS-style 2/4/5-way engine block (1x1x1 size) would be great for building a tiny probe, but could also be balanced vs. larger single-direction thrusters by being less efficient per-mass and per-volume...
     
  13. halipatsui

    halipatsui Senior Engineer

    Messages:
    1,253

    Why not weld the ship from one piece in the first place?
     
  14. Ronin1973

    Ronin1973 Master Engineer

    Messages:
    4,964
    It's not the one rope. It's when it becomes a this massively spammed device that suddenly becomes a quasi building component and we go from having rigid structures to having dynamic tree-houses built everywhere where every other block is connected by a rope.

    Then come the artificial limits placed on ropes in order to counter the server killing spamming of them.
     
  15. Malware

    Malware Master Engineer

    Messages:
    9,867
     
  16. ViroMan

    ViroMan Senior Engineer

    Messages:
    1,123
    Ropes could be implemented in a no strength way. As in it cannot hold together or pull anything. Then the server can dump the physics on the player. The server only has to keep track of the fact that the length of the "rope" can't exceed X distance before break.

    This implementation gives you power cables to power things remotely on planets and asteroids.
     
  17. PLPM

    PLPM Junior Engineer

    Messages:
    850
    Well, that´s just abusing the rope mechanics.

    You can abuse anything and make the game lag if you spam it hard enough.
     
  18. Ronin1973

    Ronin1973 Master Engineer

    Messages:
    4,964
    Yes, but as Malware stated, rope mechanics aren't fundamental to game play. So consideration of such changes, expending developer time, testing, and balancing to the game sinks to a very low priority and should.

    Working ropes versus working pistons and rotors? I'd rather have those blocks perfected as they are more useful to the game. I can't push anything with a rope.
     
    • Agree Agree x 1
  19. Sinbad

    Sinbad Senior Engineer

    Messages:
    2,788


    If we had rope, i would make these...

    Then i would buy a new cpu.
     
    • Like Like x 1
  20. Ronin1973

    Ronin1973 Master Engineer

    Messages:
    4,964
    The string is also being PULLED as fast as it is being pushed. Try that with a piece of string that's not in a loop.

    You and your string theories. :)
     
    • Like Like x 1
  21. ViroMan

    ViroMan Senior Engineer

    Messages:
    1,123
    I am seriously intrigued by how the string reacts to his prodding.
     
Thread Status:
This last post in this thread was made more than 31 days old.