Jump to content

[WIP] Infernal Robotics - Next


Rudolf Meier

Recommended Posts

57 minutes ago, Rudolf Meier said:

oh yes... I made a 3 joint test and ... I think I will keep this in the code. Seems to be good.

That sounds promising. i can think of 3 obvious uses ,  for parts i already have and  result in a much nicer thing to use.

For clarity, mine, I have to ask, as i've missed a good bit over the past few weeks, there are no part changes?  The parts are currently the same as was in the final release of IR? 

The only conclusion is that there's something off with the set up I'm using, should the included  IR parts be the same.  Will obviously test again with the latest patch and then start messing with hierarchies until it works.

 

 

Link to comment
Share on other sites

3 minutes ago, SpannerMonkey(smce) said:

For clarity, mine, I have to ask, as i've missed a good bit over the past few weeks, there are no part changes?  The parts are currently the same as was in the final release of IR? 

Yes, except for things that are not moving, like the landing foot or magnetic endeffector and things like that. The servos should be in and all (except 2 of the legacy ones that probably cannot be fixed because of model problems) should work.

Link to comment
Share on other sites

8 minutes ago, SpannerMonkey(smce) said:

which two out of interest ?

the VTOL something and the hinge that starts at 90°... I think the VTOL thing isn't very interesting and you could use the hinge that starts at 0° and rotate it to 90° to get the same result... so, not a big problem :) 

Edited by Rudolf Meier
Link to comment
Share on other sites

Decided to give this a test on a nuclear tug - engines on top, robotics extend them away from whatever they were tugging, all ready to go somewhere.

I had some... interesting results! :D

 

Edited by Rohaq
Link to comment
Share on other sites

4 hours ago, Rohaq said:

Decided to give this a test on a nuclear tug - engines on top, robotics extend them away from whatever they were tugging, all ready to go somewhere.

I had some... interesting results! :D

Check out your autostruts. Such thing is nothing new, usually happened due to autostruting on IR parts where you didn't want to be. It is hard to see craft configuration from video, try to make some screenshots from VAB, or share craft file with minimum parts from other mods involved so it can be examined what is going on if it is not case with unwanted autostruts.

Link to comment
Share on other sites

When you said 3 joint test, did you mean a single part with multiple servos built in?  As in, a robotic arm that could be all one part, instead of requiring assembly?

If so, can you post a module example so I can give it a try?  I tried putting two copies of ModuleIRServo_v3 in a single config file to try to move two different segments of a parented tree of meshes... and...  it did not work as expected.  Multiple transforms were being rotated ,not just the one I'd named as the moving mesh.

Link to comment
Share on other sites

Regarding the NaN problem and the ship teleport.

I noticed that before it occur, all servo doesn't respond to commands and they slowly drift to middle position (apparently).

Also, they stop drifting in timewarp.

 

This "servo drift" doesn't always occur, it look like it depends on the servo attached parts, and it's location in the craft. One part on the top of the craft make the servo drift fairly fast (picture in spoiler below, I'm talking about the MKS round tank), but if the part is near the bottom, the attached servo drift really slowly.

Spoiler

LlqnVBJ.jpg

All drifting servo doesn't respond but still consume EC on activation. All non-drifting servo work well.

I start wondering this is part weight related ...

 

Edit:

From the following test:

Spoiler

dJSPVyU.jpg

I used MKS container parts for their ability to change their weight alot. Red one is really heavy (2T dry + 120T content). All other are empty and just 2T.

Excepting servo C (on top of the HECS) which work well, all other servo are drifting. Note how "Cr"(ear) and "Cf"(ront) drift at the exact same speed. Those are the servo attached on the red container

All other drift are not consistent. I expected them to depend on their distance from the COM but this is not relevant.

 

Two interesting points:

  • servo stop drifting with lock engaged.
  • servo drift speed is "reset" when returning to normal simulation speed from timewarp

 

Edit 2:

Simple setup. FL-T400 Fuel tank -> HECS -> Mainsail engine. Then attach a "Joint Pivotron - Half Basic" on the side of the fuel tank.

From SPH:

  • set the fuel tank content to (or below) 144 LF/176 Ox (total part mass < 2T) the servo is not drifting.
  • set the fuel tank content to (or above) 162 LF / 198 Ox (total part mass > 2T) the servo is drifting.

I've added another FL-T400 Fuel Tank (full fuel so part mass > 2T) on the other side of the Pivrotron, it look like it has no effect on the drift.

 

As pointed by @Getsome2030 , there is some strange things with attach node. I also noted that the attached side of the pivotron is not the same when I surface attach and when I node attach. I'll play a bit with that and get back. (Edit: completely irrelevant here :( )

Edited by Gurki
Link to comment
Share on other sites

7 hours ago, kcs123 said:

Check out your autostruts. Such thing is nothing new, usually happened due to autostruting on IR parts where you didn't want to be. It is hard to see craft configuration from video, try to make some screenshots from VAB, or share craft file with minimum parts from other mods involved so it can be examined what is going on if it is not case with unwanted autostruts.

Yes, you should test that ... or try to show the autostrut connections. The problem could be a bug in KSP (how autostruts are built). I've reported that, but didn't get an answer so far.

4 hours ago, artwhaley said:

When you said 3 joint test, did you mean a single part with multiple servos built in?  As in, a robotic arm that could be all one part, instead of requiring assembly?

no, it's just a technique that is internally used to make translational joints stronger so that they remain more stable on the moving axis (e.g. an extendatron doesn't bend that much to the side and the moving part of a rail doesn't move away from the rails)

Link to comment
Share on other sites

8 hours ago, kcs123 said:

Check out your autostruts. Such thing is nothing new, usually happened due to autostruting on IR parts where you didn't want to be. It is hard to see craft configuration from video, try to make some screenshots from VAB, or share craft file with minimum parts from other mods involved so it can be examined what is going on if it is not case with unwanted autostruts.

No way, I'm taking this thing to Eeloo!

Yeah, it turned out to be an autostrut on the nuclear engine to its grandparent - the last robotics joint. Unfortunately, even when it was working without spin, it had trouble executing maneuvers with any efficiency for some reason, despite all engines being pointed correctly. Ah well.

Link to comment
Share on other sites

Finally, after recompiling (csc.exe in command line) the mod from the master branch, I no longer have any problem with the servo drift.

I don't know why but the compiler consider that every reference to "Controller" (singleton designed class into InfernalRobotics_v3.Command namespace) would point to a definition located somewhere in the .NET core library ... I fixed that by explicitly using "InfernalRobotics_v3.Command.Controller" instead of just "Controller" ...

But actually, I lost the IR filter icon in VAB/SPH :/ But not a big deal in my career actually ...

Link to comment
Share on other sites

3 minutes ago, Gurki said:

Finally, after recompiling (csc.exe in command line) the mod from the master branch, I no longer have any problem with the servo drift.

I don't know why but the compiler consider that every reference to "Controller" (singleton designed class into InfernalRobotics_v3.Command namespace) would point to a definition located somewhere in the .NET core library ... I fixed that by explicitly using "InfernalRobotics_v3.Command.Controller" instead of just "Controller" ...

But actually, I lost the IR filter icon in VAB/SPH :/ But not a big deal in my career actually ...

*hmm* ... interesting

see, they said "with .net there will be no dll hell anymore" and now we are dealing with problems like those... on top of the bugs we write ourself :) ... it's difficult to find problems like these

what system do you use? ...

I didn't try what you described (had no time) but I will... I'm wondering if that happens on my machine too

Link to comment
Share on other sites

Win7 pro here.

Regarding the compiler, I use the one from msbuild 15.0 with required options to tell him to forgot shipped configuration and compile against what's available in the KSP_64_Data/Managed folder (in an ugly way BTW ... but it worked for all mods I recompiled so ...)

Why did I use the cmdline and no IDE ? I didn't find a way to get my VS2017 working smooth ... and latest versions of mono required to be compiled from scratch using ... VS :mad:

Link to comment
Share on other sites

3 minutes ago, Gurki said:

Why did I use the cmdline and no IDE ? I didn't find a way to get my VS2017 working smooth ... and latest versions of mono required to be compiled from scratch using ... VS :mad:

That's why I'm still using VS 2010 with .net 3.5 target for all KSP mods. I recently found out how to compile for .net 2.0 with VS 2017 (we need this for MMC support on Windows 7) ... but, still I didn't try to do KSP mods with VS 2017... yet... maybe in the future, but for the moment it seems to work. No idea if there are better settings... (normally I'm working with C/C++ on my day job ... with the exception for some parts that need to be done with .net).

Link to comment
Share on other sites

I got an interesting bug using the core, struts, and Tweakscale. I only just now started using this version so I started with a simple 4 legged walker (mostly because I wanted to try sequencer only to realize just now I never downloaded it). The legs folded flat and everything vanished. I went to the space center to find this (it is not night btw). https://docs.google.com/drawings/d/1F2_uVYQkgsGLsTdsizfqiNdL1ahd1ik0SNQUNxgSKqw/edit?usp=sharing

Link to comment
Share on other sites

13 minutes ago, Mcglin250 said:

I got an interesting bug using the core, struts, and Tweakscale. I only just now started using this version so I started with a simple 4 legged walker (mostly because I wanted to try sequencer only to realize just now I never downloaded it). The legs folded flat and everything vanished. I went to the space center to find this (it is not night btw). https://docs.google.com/drawings/d/1F2_uVYQkgsGLsTdsizfqiNdL1ahd1ik0SNQUNxgSKqw/edit?usp=sharing

yeah... KSP is reacting with a destruction of the sun in such situation... or maybe Kerbin is removed from its starsystem... whatever... can you explain how you did build the craft? what parts did you use? maybe a picture of it or a save file would be helpful.

Link to comment
Share on other sites

24 minutes ago, Rudolf Meier said:

yeah... KSP is reacting with a destruction of the sun in such situation... or maybe Kerbin is removed from its starsystem... whatever... can you explain how you did build the craft? what parts did you use? maybe a picture of it or a save file would be helpful.

https://docs.google.com/drawings/d/1E7N-rQ3pWdQyWJVp_ZJjXYP_ADcoaP5empMA1vcLVkY/edit?usp=sharing

And in the tracking station there is nothing there as well.

Link to comment
Share on other sites

12 hours ago, Rudolf Meier said:

no, it's just a technique that is internally used to make translational joints stronger so that they remain more stable on the moving axis (e.g. an extendatron doesn't bend that much to the side and the moving part of a rail doesn't move away from the rails)

Ahh!  That makes sense.   Does it seem like it ought to be possible to have more than one controlled joint in a single part?   I'm working on a pack of arms that will be compatible with Sirkut's arm plugin, but I thought it would be cool to make it work with IR too... if it were possible.  I suppose I could cut the models up and release all of the different pieces, but that's not really what I'm going for with the part pack.  I want to just give users a handful of arms that 'just attach and work.'  
 

 

Link to comment
Share on other sites

Hi everyone,

I'm having some weird glitches with this mod. Here are some screenshots of a simple craft: https://www.dropbox.com/s/dg6x9mcvcez8v18/Kerbal_IR.3.0.jpg?dl=0

  • It works fine in the editor
  • It loads ok on the launch-pad
  • If I start the servo it seems to apply massive amounts of torque. Bending the whole craft and even the launch stability enhancer. Without LSA the whole craft is spinning extremely fast and starts flying off
  • happens with all other servos I tested. They all glith out in some way or another

I have advanced tweakables/autostruts disabled and I'm using a few other mods (MechJeb, CommunityTechTree, KSP Interstellar Extended, DockingPortAlignment, Maneuver Node Evolved, ScanSat) also KerbalJointReinforcement (improved version from this thread).
Anything more I can test/provide to help trouble shooting just tell me :)

 

Link to comment
Share on other sites

8 hours ago, artwhaley said:

Does it seem like it ought to be possible to have more than one controlled joint in a single part?

No, that's not possible. Because the game engine (unity) is simulating physical relations only between objects. If we would start building parts that contain multiple parts, then those are no objects anymore (for unity) and we would have to do the physics simulation ourself. I would only do this for parts that don't need physics...

6 hours ago, Sebas-chan said:

I made a robotic arm with this mod and it crashed the game. Any ideas?

yes... normally they say "no logs, no support"... that's not how it works here (because those game crashes will most likely not be in the logs), but I need some information about the craft you built... so, either a picture or a save file of the ship would help and a description of what you did. Did it crash directly when you launched it or after some time, when you went into space... those information would help to reproduce it. Only then I can start debugging it. ... it would also help if you could build the most minimalistic craft that shows the problem. That makes the job easier :) ...

Link to comment
Share on other sites

58 minutes ago, Rudolf Meier said:

No, that's not possible. Because the game engine (unity) is simulating physical relations only between objects. If we would start building parts that contain multiple parts, then those are no objects anymore (for unity) and we would have to do the physics simulation ourself. I would only do this for parts that don't need physics...

yes... normally they say "no logs, no support"... that's not how it works here (because those game crashes will most likely not be in the logs), but I need some information about the craft you built... so, either a picture or a save file of the ship would help and a description of what you did. Did it crash directly when you launched it or after some time, when you went into space... those information would help to reproduce it. Only then I can start debugging it. ... it would also help if you could build the most minimalistic craft that shows the problem. That makes the job easier :) ...

49da4f4815eb40ef24b96363394bad64.png

This is a picture of my ship. I tried to attach the arm inside the shuttle. Basically, when I launched it, it gave me the kraken. I launched it for the second time and it crashed the game. It doesn't crash when it's not attached to an probe controller

Edited by Sebas-chan
Not enough info
Link to comment
Share on other sites

9 hours ago, artwhaley said:

I want to just give users a handful of arms that 'just attach and work.'  

I'm not sure if I like to have such parts in game. It is more than a half of fun to have multiple "lego" parts that you can combine and discover for yourself if something work properly or not, discover flaws in your own designs and how to counterpart those flaws. I understand that there always will be users who would want something that is ready to plug in and play (or pray) with it.

Link to comment
Share on other sites

19 hours ago, Gurki said:

Finally, after recompiling (csc.exe in command line) the mod from the master branch, I no longer have any problem with the servo drift.

I don't know why but the compiler consider that every reference to "Controller" (singleton designed class into InfernalRobotics_v3.Command namespace) would point to a definition located somewhere in the .NET core library ... I fixed that by explicitly using "InfernalRobotics_v3.Command.Controller" instead of just "Controller" ...

But actually, I lost the IR filter icon in VAB/SPH :/ But not a big deal in my career actually ...

how did it "drift" ... I cannot see the problem

and how did you connect the parts? Flt-400, then on it's bottom the hecs and then again one down the Mainsail? what was the first part? the flt400?

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...