Welcome to Keen Software House Forums! Log in or Sign up to interact with the KSH community.
  1. You are currently browsing our forum as a guest. Create your own forum account to access all forum functionality.

What is -actually- being done about rotors and pistons?

Discussion in 'General' started by Zathura, Jul 28, 2016.

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

    Messages:
    1,750
    @StreamBG I can tell you right now how to temporarily fix the drill rig at the end (with all the pistons) Just for the sake of argument remove one set of the two pistons on each stack (you may wish to rebuild them in the middle, but for test purposes just removing one should fix the spins
     
  2. M3A7 Trainee Engineer

    Messages:
    68
    Are we actually finally acknowledging that Keen isn't capable of fixing most of the problems they have made? My my my.
     
    • Like Like x 2
  3. RageMasterUK Apprentice Engineer

    Messages:
    201
    We really deserve a fix to these things. Or at least some sort of official comment. Rotors and pistons are the most interesting blocks, but the most unstable. Just having them in the G menu means I have to try and use them on cool things!

    My mammoth tank mobile base large-grid NEEDS a rotor/piston folding ramp, but every time I make it and reload my world it self destructs. Ive rebuilt it a dozen times, each time blowing up. Its disheartening. Keen if you are lisening please....

    It wouldnt be so bad except that rotors and pistons are what sets this game apart from say, Empyrion, as a space-building game. We need these blocks to function right, or tbh theres no real selling point from this game over some of its competitors IMHO.

    I eagerly await their fix, in the meantime all my advanced builds need to be put on hold and I should play something else.
     
    • Agree Agree x 1
  4. Vigo the Dudepathian Trainee Engineer

    Messages:
    87
    I think the official word is they've working on it.
     
    • Like Like x 1
    • Agree Agree x 1
  5. RageMasterUK Apprentice Engineer

    Messages:
    201
    I hate to go on, but I can't actually remember when they were this bad. The instability seems to be on save/load right now. I can build my rotating sliding door thingy and it'll work just fine in the instance of the game I built it in, but then 95% of the time when I reload theres catastrophic failure.

    I'd be happy to post videos to highlight my specific plight, if it'd help you out Keen! :D
     
    • Agree Agree x 2
  6. GrindyGears Senior Engineer

    Messages:
    1,750
    So, i did some testing just now, and it seems they have restored some functionality to rotors and pistons, in my tests i was once again able to:

    -place more than 2 rotors connected to a single grid (in this case 4 as is my usual amount) without the massive mysterious force.
    -Place multiple pistons that are all attached to one grid i only tired with two but its behavior changed greatly from last patch, this is good.
    -didnt try big to small conversion kits, but when im not on my laptop i'll try.

    I also observed an interesting behavior in the copy paste system.

    when i had a rotored or pistoned device ( it was combo so i dont know about individual behavior) when i directly used the COPY function (ctrl+C) when i placed the object it appeared as if it was trying to snap to the parent/original grid almost as if it didn't yet recognize that it was a separate/new grid which lead to it clipping which obviously caused a kaboom. However when using the DELETE function (ctrl+X) and pasting i had absolutely 0 issues, and no amount of additional copies gave it any trouble. I'm not enough of a code monkey to really look into this but i'm going to loop @JoeTheDestroyer in on this, because he found another issue in the pistons ( i think) where it wasnt properly updating where the piston should actually be on paste/spawning

    Blueprints that previously were violent detonations are now not nearly as implosive (tested on my excavator and the only thing that exploded was the janky small to large to small grid rotor i was using (4 point connectors causing conflict i think) which is an improvement from last patch, however when i spawned my drill rig it exploded, so obviously still room for improvement.

    On a tangent note: I think one thing that would solve some of the issues plaguing pistons would be to give us more options on the length of them. if the game has trouble calcuating 4 stacked pistons (8 blocks) why not just give us a variant thats 8 blocks long? it gives same overall length at 1/4 the connection points
     
    • Agree Agree x 1
  7. Lothos Apprentice Engineer

    Messages:
    312
    one of the reasons i use modded pistons is to have pistons of varying lengths to make use of only one piston per application.

    Frankly, without rotors/pistons to use there isn't much "engineering" going on in the game :/
     
    • Agree Agree x 1
  8. Harrekin Master Engineer

    Messages:
    3,077
    Even Lego has moving parts...
     
    • Like Like x 1
  9. Spets Master Engineer

    Messages:
    3,136
    somebody expected that PN mechanics blocks was going to be free of clang?

     
    • Funny Funny x 4
    • Agree Agree x 2
    • Like Like x 1
  10. DDP-158 Master Engineer

    Messages:
    3,748
    WhaaaaaĆ aat?!?!?! You mean to say this is a recurring issue that expands beyond one company? All this time I was led to believe by the armchair developers that this was a issue that could be fixed in 30 seconds and that the devs just kept it around to annoy customers
     
    • Funny Funny x 3
    • Like Like x 1
    • Agree Agree x 1
  11. Harrekin Master Engineer

    Messages:
    3,077
    A fine tribute to Lord Clang.

    His blessings I grant to its builder this day.
     
    • Agree Agree x 2
  12. JohnDoe Trainee Engineer

    Messages:
    58
    Nobody's ever said other games don't have clang, or that it was an easy problem. Just that they A) have to solve it and B) need to show some progress after all this time.

    We'll just have to see if it's unsolved two years from now in PN.
     
  13. Thedevistator Senior Engineer

    Messages:
    1,942
    Also worth noting Unity uses PhysX not havok so changing the engine isn't neccesarily the solution either.
     
    • Agree Agree x 4
    • Funny Funny x 1
  14. May Rears Apprentice Engineer

    Messages:
    225
    Oh the new shenanigans now that with the latest update (Dev). You can control thrusters on subgrids, General Clang has a whole new bunch of ways of blowing us up :)
     
    • Funny Funny x 1
  15. cPho Trainee Engineer

    Messages:
    14
    Aha. Ok here I was thinking I'd done something wrong.

    On my missile rig I have connected pistons together and when I pasted it in, it would spin endlessly with such violence that it blew LARGE WARHEAD sized holes in the ground where it impacted. Like wow that is powerful for a small grid of this mass.

    So when I manage to paste it just close enough to the ground so that the landing gear bites and holds it still, I see that one of the pistons is twisted about 30 degrees. By activating the pistons and then reversing them for a split second it corrects itself, putting to rest my fear that somehow I'd attached it at 90 degrees to the other one or something.

    Because when I was trying to connect two pistons to a single grid using merge blocks, this same top piston would rotate Part A of the joint platform at 90 degrees to where I had built it... but this could be fixed by deleting that Part A and rebuilding it in the correct orientation, and then connecting it with the merge block to part B, then welding them together and deleting the merge.

    BUT, now that I paste it in it does that nuclear spin thing right. So I delete that piston and rebuild the entire thing again, taking a new copy and blueprinting it. And when I paste THAT in, the entire thing detonates with so much force that I found large chunks of it almost 2000m away. Seriously, I have to find a way to weaponise this!

    Long story short I thought I was doing something wrong till I read this thread, so thanks, I won't be quite as disappointed now. No wonder it was suggested that I remove as many pistons from my grids as possible.

    EDIT > I took some pics of the newly pasted in rig to show what I'm talking about.

    [​IMG]
    When I paste it close enough to the ground and rotated just right, the landing gear locks and it doesn't spin like some sort of satanic baby toy. But notice the pistons... twisted and warped. I didn't save them that way. I saved it in a format AFTER I'd corrected this.

    BUT:
    [​IMG]
    If I reverse then reverse again the pistons (so make them extend for a split second, then back to rest position... like double tap on it) the pistons correct themselves. Problem solved.

    So a thought:
    Maybe a grid using lots of pistons could be set up to autodock a gear to the ground to pin Klang's hands behind his back, and then a timer block (or manual if yer patient) activates all pistons for a split second, and that's the saved blueprint? Then release or grind off the gear and yer blueprint is ready to go...? I'ma try that at some point when I'm not focused on making the missile rig itself function. Thought I'd share in case someone else has other thoughts or wants to experiment.
     
    Last edited: May 11, 2017
    • Like Like x 1
  16. GrindyGears Senior Engineer

    Messages:
    1,750
    Currently there is a bug caused by grid orientations not matching when you merge them that causes those nasty twists in your pistons, a couple of patches ago it caused the attached grid to twist in an unusual way.

    You can get around this by being very mindful of the pistons "grid pivot" (under info tab check show grid pivot) As long as all of the pistons match as well as the grid they may be merged to are the same orientation, things should be fine.

    Unfortunately I have yet to find a way to counteract the spinning effect from pistons, basically all of my stuff is heavy enough on a planet that the spin doesn't bother it, and when constructing it in space, it's always landing geared to a station.
     
    • Informative Informative x 3
    • Agree Agree x 1
  17. cPho Trainee Engineer

    Messages:
    14
    That is EXTREMELY useful informationzzz.

    I rebuilt the piston connections by paying attention to the grid pivot that they inherited... I noticed that the grid pivot point for my other pistons that didn't have "twist" issues were (by coincidence) already aligned... I also notice that some other points in the grid are not aligned but they're not giving me issues. I just plopped piston, bad grid, delete rotate, plop, bad grid, delete rotate, plop, good grid, next piston... worked nicely.

    After aligning the pistons, reconnecting with merge block, and copy-pasting the grid into the world again, zero spin. Just sinks onto the gear and locks in place. Thanks a bunch GG.
     
    • Like Like x 2
  18. The Churrosaur Junior Engineer

    Messages:
    739
    @damoran I don't know if this has been mentioned yet, but check this out. Its a fully vanilla ingame program allows you to control VTOL thrusters on rotors. It's as amazing as it sounds
     
  19. Devon_v Senior Engineer

    Messages:
    1,602
    As far as I understand the issue is simply that Havok and VRAGE get into a feedback loop of a sorts where Havok says "well, this moved there" and VRAGE says "well, then that happened", so Havok goes "well then this had to move again", so VRAGE says "well, then that happens again" Little phantom movements, rounding errors, stuff that is difficult to plan for. You can't just apply pure game logic to it because the whole point of the physics engine is to allow completely emergent behavior. It takes tons of sanity checks to try to keep the reactions "in bounds" as it were without preventing things from exploding when that's what the player wants to hapoen. And the more checking and calculations you do, the slower things get.

    There's a mod on the workshop which Keen is aware of, which adjusts the physics to be a bit more "arcade" so that minor issues are effectively ignored. It makes the physics less accurate, but it makes stuff far less explody.

    Basically they have to figure out how to make Havok work wrong in the right way to keep the game fun.
     
  20. Hotshot Jimmy Senior Engineer

    Messages:
    1,404
    I think it is always going to be a balancing act. I doubt we'll ever have what we have now and it be stable too but I'm sure there is a good point where it still feels like "physics" are being involved but not at the detrement of gameplay. They'll get it one day but as has been said above it is something ALL physics based games have issues with.
     
  21. GrindyGears Senior Engineer

    Messages:
    1,750
    Well, this is a bit of a self promotion, and also a bit of a "to be fair to keen"

    They have done quite well imo, to support even large physics creations. I mean I've done some nutty stuff in my time and this is probably near the top:

    [​IMG]

    This is the older version (I have a finished/decorated version, but no pictures like this)

    Its like a 600 tonne 100m + reaching small grid excavator with large grid pistons, riding on the back of a 300 tonne small grid 8x8 tractor that uses large grid wheels and rotors to drive with a trailer that can basically kneel down, and allow direct drive onto the back access...

    Now I haven't yet tested these in strictly planetary conditions with block damage on, but I have tested them separately with damage on using digis gravity mod. So I have reasonable ground to believe it would work just fine.

    I will always stick to my general moto for SE:

    The solution to clang is more often than not: more clang, if one rotor is breaking, add another, if one piston stretches too much, add another. If a small grid rotor isn't strong enough, use a large grid one, if that's not enough use gyros to help it.

    Given that we can't just make our pivot "pins" bigger make them more frequent to distribute the load.

    Of course this all comes with the caveat, that it can maintain a synchronized entity between the server and clients. If they iron out MP issues (which they have said they are close) most of Clang should go away.
     
    • Like Like x 3
    • Agree Agree x 1
  22. VigotheDudepathian Trainee Engineer

    Messages:
    31
    I have been saying for a bit keen just need to do this. Have an "arcade mode" that is immensly more forgiving to the physics and then a realism mode the keep the purist complainers happy. They will still complain about lost dev time, but at least things wont blow up all the time. Also, got a link to the mod?
     
    • Agree Agree x 1
  23. Sinbad Senior Engineer

    Messages:
    2,687
    this is now my favorite SE quote. you should use it in your sig.

    and it should be a loading screen quote?
     
    • Funny Funny x 2
    • Like Like x 1
    • Agree Agree x 1
  24. Devon_v Senior Engineer

    Messages:
    1,602
    Mobile Steam is a pain. It not only took five minutes to just to find my mod list, it won't provide direct links like the desktop version. Search for "PhysFix: A Slightly Arcade Physics Fix"
     
  25. damoran Junior Engineer

    Messages:
    601
    There's not much real about the physics in this game either way, tbh.

    Why I've been saying for some time to just have the mechanical bits work, without me having to be an expert in havoc to use rotors/pistons.

    If they insist on keeping the physics the same, could they give us ways of A) knowing the breaking stress on rotors/piston B) better ways to counter these stresses. Give me some indication that placing 50 blocks on that rotor is gonna break it without adding a gyro or another rotor. Might be a step in the right direction if they cannot just work outright.
     
    • Agree Agree x 1
  26. GrindyGears Senior Engineer

    Messages:
    1,750
    While I do agree to some extent, most of it is actually pretty straight forward, there are a few exceptions that I think I've outlined earlier in the thread. Things like the 8mm offset needed for small rotors to prevent spinning is an example of that. But it's a thing, once you have that in your mind it should be almost sub conscious.

    To split that up:

    A) they can give you numbers till you're blue in the face, but what real good is it to know? The effort required to actually calculate how much force is being applied in many cases is a pain in the ass at best, multiple center of masses, 3D space centroids, dynamic loading, inertia... Basically the number of factors that go into those calculations makes figuring out if you're breaking that threshold is awful.

    B) what exactly would you suggest as a way to better dampen the forces? They could try to add things like springs, but then you get back into calculating required forces...

    I mean, you said 50 blocks as an example so let's roll with that for now:

    Are those 50 straight out from the rotor? Or are they all clustered close to the center? There really is no single exact guideline(s), so I don't think keen could even realistically give those kind of details without some ridiculous amount of work to figure out tons of approximations. Which the user then has to Dredge through to try and see if what they built will survive.

    Most of the mechanics are just sorta learn as you go, I personally haven't had any major troubles with overloading in a long time, I've pushed rotors and pistons to upper limits in the past, you'd be quite surprised how resilient they cab actually be.

    [​IMG]

    As an example, those pistons that control the boom (first segment) are at a horrid mechanical disadvantage, with 430 tonnes on the end of a 100m lever the force is insane, and for the most part it's all passed through 6 rotor hinges for pivot points and 4 rotor hinges connecting the pistons to the grid.

    Tldr: no definite guide, learn through experience
     
    • Like Like x 2
  27. CrazyEd Trainee Engineer

    Messages:
    77
    Does this constitute an SE horror story ?
    [​IMG]
    In this screenshot there are two medium size solar (tracking) arrays in a U configuration (3 rotors each). A refueling / load/offload crane comprising a rotor and three pistons with a connector on the end which services the vehicle (a ten wheeled chap, which has 4 piston jacks) and the hydrogen powered shuttle in the lower right corner.
    Framed in the cranes pistons is a small builder ship which has two arms with a welder on one end and a grinder on the other, it uses rotors to swap between the tools. It's biggest problem to date appears to be a mass calculation problem caused by the sub grids (i'm guessing) as it would just sink to the ground inspite of having enough thrust. This was solved by adding thrusters to the arms, that get turned on or off depending on the arms orientation.

    What can't be seen in the screenshot is a vertical gate that is driven up out the ground by two pistons, a satellite with a small solar (tracking) array about 45km above the base which provides a com link to a ore processing facility on an asteroid a bit further away. That runs almost soley on battery / solar having 4 large tracking arrays.
    I did have a problem with the one array where the one rotor failed to merge properly with the main arm which lead to continual damage to the arm. It was corrected and seems ok now

    Am I tempting fate/Clang?
     
    • Funny Funny x 1
  28. damoran Junior Engineer

    Messages:
    601
    Yea nobody is going to do that, I was thinking more of a visual system.

    Doesn't really matter what you build on the rotor/pison, I was just suggesting a way to tell if it is going to break because of what I have attached to the rotor head. I mean it's gotta know already right? Whatever Havoc is doing to figure out the result of how the rotor works in real time. There seems to be a tipping point where the rotor just breaks or goes unstable but many times we don't know until we try to move something or move the ship.

    I'm thinking a Green/Red light, somewhere on the rotor base.

    Green is good, stress currently placed on rotor/piston is acceptable
    Red is beyond the rotors capabilities

    I don't know enough about how they work to suggest intelligent dampening techniques other than simply having the rotor connected sub grids ignore their individual physics calculations and have the game apply the calculations from the main grid instead. I'm sure there's a million issues with doing that, just the first thing that comes to mind.

    But yeah I am a super noob with rotors/pistons compared to what you've done with these parts Grindy, I'm just trying to figure out ways to make rotors/pistons easier to fiddle with.
     
  29. GrindyGears Senior Engineer

    Messages:
    1,750
    Unfortunately that is not always the case, something that is super massive and heavy might put an enormous force on the rotor, but depending on how that force is applied it makes a major difference.

    I have things that can rotate safely and without issue at 3-9 rpm, but if you ramp the speed up to say 30 rpm it will destroy itself.

    What you're talking about is usually found in a year two mechanical engineering course at universities. Generally referred to as "dynamics" it's the study of forces and acceleration(s) undergone by various machines, mechanisms, and rigid bodies.

    Static force calculations vs dynamic load calculations often yield extremely different results, with the latter being much higher due to things like inertia. What might be just fine by your green light, might instantly turn red as soon as it starts to move.

    Again, I think the best way to learn your breaking thresholds is to just build stuff and push it past breaking point. (do so incrementally so you can get a more accurate bead on the limits)
     
    • Informative Informative x 2
    • Like Like x 1
  30. PARohT Trainee Engineer

    Messages:
    28
    So basically we can conclude that still nothing is being done to help preventing clang from rotors/pistons?
     
    • Disagree Disagree x 2
Thread Status:
This last post in this thread was made more than 31 days old.