Jump to content

[1.0.x] [V1.9f] Kerbal Foundries wheels, anti-grav repulsors and tracks


lo-fi

What to work on next?  

1,282 members have voted

  1. 1. What to work on next?

    • More wheels
      123
    • More tracks
      453
    • Rover bodies
      241
    • Landing gear
      137
    • Landing legs
      108
    • Something completely different
      193


Recommended Posts

on the unity / physx news front I caught this:

unity 5.5 release target: nov 2016

- physx 3.3.3

now physx 3.3.3 might fix some of the inherent issues of wheels:

Vehicles

  • Fixed:
    • Sticky tire friction was unreliable with more than one substep but is now fixed. This defect led to vehicles sliding down slopes where the sticky friction should have held firm.
    • An error in the jounce speed calculation that led to lift force at high forward speeds has been fixed. This defect led to instabilities at high speed.
    • Improved documentation for PxVehicleSuspsensionData::mSprungMass.

We have to hope that in future ksp will upgrade to unity 5.5 and physx 3.3.3 :)

source: https://unity3d.com/unity/roadmap http://physxinfo.com/files_new/PhysX_3-3-3_release_notes.html

Link to comment
Share on other sites

  On 10/6/2016 at 3:09 PM, riocrokite said:

on the unity / physx news front I caught this:

unity 5.5 release target: nov 2016

- physx 3.3.3

now physx 3.3.3 might fix some of the inherent issues of wheels:

Vehicles

  • Fixed:
    • Sticky tire friction was unreliable with more than one substep but is now fixed. This defect led to vehicles sliding down slopes where the sticky friction should have held firm.
    • An error in the jounce speed calculation that led to lift force at high forward speeds has been fixed. This defect led to instabilities at high speed.
    • Improved documentation for PxVehicleSuspsensionData::mSprungMass.

We have to hope that in future ksp will upgrade to unity 5.5 and physx 3.3.3 :)

source: https://unity3d.com/unity/roadmap http://physxinfo.com/files_new/PhysX_3-3-3_release_notes.html

Expand  

Since things are going to south lately, i'd not depend myself on upcoming updates.

Link to comment
Share on other sites

  On 10/6/2016 at 5:37 PM, qromodynmc said:

Since things are going to south lately, i'd not depend myself on upcoming updates.

Expand  

What makes you say that?  All you know is that several employees left, probably when their contract allowed them to.  Some of them were short-term contracters brought in for this release.

In addition, the total development staff is still larger now than it was at the beginning of the year.

Link to comment
Share on other sites

  On 9/28/2016 at 7:29 AM, lo-fi said:

Edit: oh, got it now. An interesting challenge is sequencing the legs so they don't just end up lifted all at the same time :rolleyes:

Expand  

You could add an action group-compatible button to allow the player to reinitialize the leg positions to a nice ripple. Maybe even derive your information from the same place you get relative steering weights.

Link to comment
Share on other sites

  On 10/7/2016 at 11:30 PM, lo-fi said:

Hey, good to see you're still around :)

Life has once again got in the way, I'm afraid, so apologies for the lack of progress. Thank you everyone for the patience.

Expand  

Hey man, there are no deadlines in modding :) that's the wonderful thing about the "Komunity"

Link to comment
Share on other sites

  On 10/7/2016 at 11:30 PM, lo-fi said:

Hey, good to see you're still around :)

Expand  

Thanks! Finally submitted my PhD thesis so have time to work on things for KSP again :D (http://forum.kerbalspaceprogram.com/index.php?/topic/59359-wip-msis-infernal-robotics-model-rework-v02-pre-release-beta/&do=findComment&comment=2792757)

Link to comment
Share on other sites

  On 10/12/2016 at 5:16 PM, lo-fi said:

After much messing, I cant figure out what KSP doesn't to mess with the position of my joint anchors. It just doesn't behave the same as in the Unity editor... ARGH 

Expand  

And every now and then we think to ourselves " we do this for fun????!!!! " I feel; your pain, i too am being abused by 1.2s physics

Your endeavors are appreciated by many, here's hoping for better times ahead :)

Link to comment
Share on other sites

  On 10/12/2016 at 5:16 PM, lo-fi said:

After much messing, I cant figure out what KSP doesn't to mess with the position of my joint anchors. It just doesn't behave the same as in the Unity editor... ARGH 

Expand  

Hmm... that is a bit disheartening.


Well, the semi- good news is that the force-based setup did work mostly.  Could probably clean up most of its problems with a bit of careful coding / heuristics.  The 'auto-strut' feature introduced with KSP 1.2 might also help -- IIRC it was added to 1.1 as a solution to jittering in the stock wheels/code, so may be exactly what we are looking for.

Would still need to come up with a solution for bump-stop and sticky friction though; as I've not been able to find a solution to those from a force-based physics perspective.


Will try and take some time this week/weekend to re-examine the force-based setup.  You are also free to poke around with it if you want -- it is the master branch of the repository.  Should already contain a compiled .dll and a ModuleManager patch to convert all of the stock wheels to the custom plugin (in the GameData directory in the repo; should be able to drop that in your KSP folder and be good to go).  Testing might be as simple as that -- drop it into the 1.2 game, turn on the advanced-tweakables auto-strut, build and launch a craft, and play with the spring/damper sliders and see if/how it wigs out.

 

 

Link to comment
Share on other sites

  On 10/12/2016 at 5:28 PM, Shadowmage said:

The 'auto-strut' feature introduced with KSP 1.2 might also help

Expand  

If you guys decide to implement this, please, please, make it toggleable. Defaulting to heaviest is fine, just don't lock them. I've been trying to figure out how to build complex suspension nonsense since 1.1, and I can't do that with the autostruts all locked to heaviest. 

Additionally, I've been thinking about a fully active suspension controller using a PID loop. The idea is that you have an electromagnet applying force as dictated by the computer. I imagine this should work in real life, but would it work ingame? The discrete timesteps tend to screw with things like this, as you're well aware, but maybe this option would be exempt.

Link to comment
Share on other sites

  On 10/12/2016 at 5:23 PM, SpannerMonkey(smce) said:

And every now and then we think to ourselves " we do this for fun????!!!! " I feel; your pain, i too am being abused by 1.2s physics

Your endeavors are appreciated by many, here's hoping for better times ahead :)

Expand  

Dude, 1.2 looks like it's gonna make a mess for a while. A lot of potential. Just massive changes. I only hope we don't loose a lot of great talent to the horrer that is "Updating"

And I wholeheartedly agree. All of you guys are very much appreciated.

Link to comment
Share on other sites

@lo-fi

changing a topic for a moment, do you still have that particle setup from your bulldozer part? I was thinking of using it in my mobile screener unit animation (just for fun) where screened rocks are thrown outside via the conveyor belt :P

 

Edited by riocrokite
Link to comment
Share on other sites

I'd like to thank everyone for your diligent work and dedication to giving me the ability to make explosions. Oh and: 

I hope everyone has a great weekend!!!


(BTW I can't wait for the tank treads to be updated...I wanna make mobile bases.")

Link to comment
Share on other sites

Recompiled the KSPWheel master branch against the KSP 1.2 libraries.  Everything appears to be working about as it did before using the physics based suspension setup; which means it mostly does work.  Patch is included for the stock wheels for testing purposes.  For some reason though I can't find the auto-strut/advanced tweakables in my dev install, so have been unable to test if the auto-strut fixes the jittering.  Hopefully will have a bit more time later today to work on that a bit/do more testing.

Link to comment
Share on other sites

Howdy!  Software Tester reporting in!  Anything I can do to help y'all out?  I looked through your latest on Github and I *think* I grok where you're at...  Though I'm not sure what this "bump stop/sticky friction" is...  Jargon?

Link to comment
Share on other sites

  On 10/15/2016 at 8:05 PM, Salawat said:

Howdy!  Software Tester reporting in!  Anything I can do to help y'all out?  I looked through your latest on Github and I *think* I grok where you're at...  Though I'm not sure what this "bump stop/sticky friction" is...  Jargon?

Expand  

Thanks for the offer :)

Those are both terms that Lo-fi and I settled on so we knew what we were both talking about; you might not find them anywhere else.

For us --

Bump Stop = anti-punchthrough, overcompression-prevention, basically keep the wheel above the ground even when it is overloaded or heading towards the surface at ludicrous speeds.

Sticky friction = The ability for an object experiencing friction to 'stop' on an incline surface.  Commonly referred to as 'static' friction (where dynamic friction is that of moving objects) in physics.

 

 

On a bright note -- the testing of the physics based suspension went about as well as expected.  It worked.  Still had jitters at close-to-critical-damping.  Largely usable though as soon as the bump-stop gets figured out.  Sticky friction may be as simple as force re-positioning the transforms/part whenever undesired movement is detected.  Jitters may be fixed through some careful code (examine if damper is rapidly going pos, neg, pos, neg, and apply some corrective / limiting measures / reduce damp or spring force when oscillating).

Nearly finished with the 1.2 update push for my mod (few more things to fix, 6 or so more models to make), and then I should be able to dedicate quite a bit more time to finishing / fixing up the custom wheel solution.  I really would like to see KF return for my next career playthrough :)

Link to comment
Share on other sites

Still, it's cool seeing some progress getting made, even if it's minimal.  What matters most for Kerbal Foundries is having working wheel code.  Anything after that is icing on the cake.  Believe me, even I'm sitting here patiently waiting for the return of useful wheels, tracks, and other funny business.

Sure, there's ideas I've scattered about in here, but one, they rely on having the wheel code working, and two, they're ideas without implementations, and without expectations on them.  I'm comfortable hearing "Sorry, but that can't be done."  It's why I throw ideas around, because I can't do them myself.  Sure, I often config-bash stuff together and make a mess of things, but that's far from actual modding (though I will admit to feeling pretty-damn-smug when I took a wheel that used KSP 1.0.5's wheel colliders and forced it to use KF instead.  Only took me three hours), and I can't do 3D stuff.

But, thank you for continuing to work on this.  I know many people in other modding communities who would give up if their mod got broken this badly by a game update.

Link to comment
Share on other sites

  On 9/6/2016 at 4:43 AM, 0111narwhalz said:

I believe your suggestions are a bit premature. Keep in mind the guys are trying to get the physics to work with them at the moment; we'd be lucky to see a stock conversion by 1.2. Additionally, if you want these items to be present in the game, learn how to model and create one yourself! Grab yourself Blender or something and fab up a couple of models. The documentation is easy to find, and all the tools are free of cost.

Additionally, some of your suggestions are outside the scope of Kerbal Foundries (by my interpretation). Prebuilt tank hulls and a specific tank are the realm of your own mod, based upon the groundwork of Kerbal Foundries.

Expand  

Just suggestions. i do agree that prebuilt things would need a different mod, but more types of tank suspension, such as treads you can wrap around parts, would be nice.

Link to comment
Share on other sites

Guest
This topic is now closed to further replies.
×
×
  • Create New...