1.1.2 Magic Smoke Industries Infernal Robotics 2.0.2

[Drew Kerman]

hey all, I have a problem with getting the Legacy parts to show up in the VAB. I researched them back in 0.90 or earlier and with 1.0.4 they've mostly been moved to nodes further down the tree than where I was before upgrading KSP, so when loading craft files from 0.90 I can't launch them. First I tried just a MM patch to adjust their tech nodes:

@PART[dockingwasher_stdScaleable]:FINAL
{
  @TechRequired = Start
}

But that didn't work, and they didn't event show up as researched in my R&D building. So then I went into my .sfs file and added:

Tech
{
  id = start
  state = Available
  cost = 0
  ... other parts...
  part = dockingwasher.stdScaleable
}

Now they showed up:

That's the docking washer in the lower right. It no longer even appears in its original tech tree node in the R&D building. And yet...

????

[Dr_Goddard]

I am having a hard time with the docking washer...what can and cannot dock to it...can two docking washers connect? sighhh....

Docking washers are just swivel joints. They are not docking ports.

[harpwner]

any chance of making toggleable presets? I'm making these cool cargo bay openings for my helicopter and jet for missiles... but I have to assign next and previous presets in the action groups.

[Motokid600]

What exactly happens when IR parts are attached to docking ports? Im looking to have a probe core on a rotator joint attached to a docking port so the rest of the ship can spin, but i get level avionics.

[allmhuran]

What exactly happens when IR parts are attached to docking ports? Im looking to have a probe core on a rotator joint attached to a docking port so the rest of the ship can spin, but i get level avionics.

Not sure why that would require the use of a docking port, but in any case, you may not get the behaviour you're expecting. Do you want to use the rotator to rotate the ship, or are you going to rotate the ship with normal flight controls and use the rotator to joint to simultaneously orient the probe? I expect the former, because the latter would require that you have exactly the right amount of speed and acceleration on the servo to negate the (variable) roll speed of the ship.

But if that's the case, what will actually happen when you tell the rotator to move is that the probe will rotate (instead of the ship), because it weighs a lot less than the rest of the ship. If you want to keep the probe level while the IR joint rotates the ship you are going to need a hell of a lot of control authority on the side of the joint that has the probe on it.... ie, a bunch of reaction wheels, or probably RCS.

[allmhuran]

Not sure why that would require the use of a docking port, but in any case, you may not get the behaviour you're expecting. Do you want to use the rotator to rotate the ship, or are you going to rotate the ship with normal flight controls and use the rotator to joint to simultaneously orient the probe? I expect the former, because the latter would require that you have exactly the right amount of speed and acceleration on the servo to negate the (variable) roll speed of the ship.

But if that's the case, what will actually happen when you tell the rotator to move is that the probe will rotate (instead of the ship), because it weighs a lot less than the rest of the ship. If you want to keep the probe level while the IR joint rotates the ship you are going to need a hell of a lot of control authority on the side of the joint that has the probe on it.... ie, a bunch of reaction wheels, or probably RCS.

PS: You *could* accomplish this with KOS, ie, accept that you're rotating the probe, and use KOS to continously command the rotator joint to maintain the probe's orientation with a basic PID controller. OR (much easier but less accurate) you could use an uncontrolled rotator joint and turn on SAS, working on the assumption that the SAS of the probe is enough to maintain its orientation. Just don't have any reaction wheels on the rest of the craft.

Edited August 19, 2015 by allmhuran

[Dunbaratu]

Not sure why that would require the use of a docking port

Not the original poster here, but I'm guessing probably because the stock game doesn't let you just pick "control from here" on any part you feel like. Getting the navball to show the orientation from the point of view of the end of the IR arm is probably the goal, and sticking a docking port on the end of the arm and clicking "control from here" on it would be one way to try to do that.

[Motokid600]

The only reason it needs to actually go on a docking port is because I failed to incorporate it into the design earlier. But what I'm thinking is vastly simple. When I say rotator joint I don't mean motorized. A free spinning joint with a reaction wheel and a probe core. I spin up the entire ship. Then have SmartAss hold the probe level with that reaction wheel.

It was a last minute decision to spin the craft. If I'm going to do this I'd like to have level avionics. But unfortunately... I just realized that every time I time warp I'd have to spin the ship up again. Bah. Oh well.

[allmhuran]

Indeed so.

But yes, with the power of mechjeb compelling the probe to stay level, that design would probably work just fine. The issue with IR and docking ports is only around what happens when you dock craft together and the game recalculates the part tree, which can mess up IR joint orientations. I've never actually encountered a problem from attaching a docking port directly to an IR joint (if no docking was actually happening), but if you're worried about that specific wording from the OP, just hide a cubic strut between them.

I do wonder, though, what should happen when you provide roll control input to the ship

[Motokid600]

Indeed so.

But yes, with the power of mechjeb compelling the probe to stay level, that design would probably work just fine. The issue with IR and docking ports is only around what happens when you dock craft together and the game recalculates the part tree, which can mess up IR joint orientations. I've never actually encountered a problem from attaching a docking port directly to an IR joint (if no docking was actually happening), but if you're worried about that specific wording from the OP, just hide a cubic strut between them.

I do wonder, though, what should happen when you provide roll control input to the ship

Thanks i might give this a try with that persistent rotation mod. Roll control? Well the ship will already be rolling. The leveled control point on the IR part will just roll If want and it will go off level. However its yaw and pitch im worried about. Or not worried about because i have a theory. If you notice that when it comes to RCS thrusters KSP will activate the thruster that is closet to the axis your inputing control too. So if my ship is spinning and i want to pitch up. As the rcs thrusters spinning on the main ship come into the axis upon which im pitching they will activate, deactivate, activate, deactivate as the ship spins. Might work. IIRC thats how alot of real life space craft who use spin stabilization ( Juno for instance ) yaw and pitch correctly. The thruster thats rolling around fires at the correct axis. And then the next one and so on.

[Myslius]

VTOL Rotor Mk1 is too flexible. Is there a way to remove or decrease the flexibility? I've tried to use this rotor with nukes, nukes have really long burn time (40 mins on my rocket) so i need to use phys time-warp to decrease that travel time, when i do that... the rotor bends a lot (30-40 degree angle).

[Ziw]

VTOL Rotor Mk1 is too flexible. Is there a way to remove or decrease the flexibility? I've tried to use this rotor with nukes, nukes have really long burn time (40 mins on my rocket) so i need to use phys time-warp to decrease that travel time, when i do that... the rotor bends a lot (30-40 degree angle).

Unfortunately Legacy parts are a bit wonky. Could you try using Zodius excellent Reworked parts? They don't have the exact model for VTOL rotor, but you can construct one yourself using a combination of parts (rotatron + mount part of your choice).

[allmhuran]

I have found I'm having a new issue, never seen this one before but I haven't been working on my robotics for a while, been doing geometry. As such this may have come with the latest update.

It seems like my servos are snapping to limits when they get within a couple of degrees of a limit that has been set. A short video demo is included below.

This only occurs when two conditions are true:

1) The servo speed is set very low, and

2) The acceleration value is set very high.

The values at which the problem starts becoming visible are around servo speed about 0.2 or less and acceleration greater than 30.

To me it looks a heck of a lot like a rounding issue with acceleration.

https://youtu.be/55-jt7ic1qY

Possible solutions if the rounding cannot be fixed due to precision limits? I dunno... maybe don't invoke acceleration when reaching a limit, only use it when a player stops a move command. Or allow an option to use the old system without acceleration (acceleration really messes up my craft since I require servos to reach positions at set times).

[Ziw]

I'll look into the code, this snapping is really odd.

[allmhuran]

Roger, I can sorta work around it in the meantime.

Getting close to done but I guess I should wait until 1.1 now because I'm sure all my stuff (and yours) will break

The only things I can't work around are the complexities about when commands (move, stop, etc) take effect. I've got everything working, but it does require that I issue a stop command to all servos and issue the move command at least one frame later. This introduces a tiny pause in the movement that makes it every so slightly jerky, but it's nothing too serioius.

Current state of work (picture only)

http://imgur.com/gallery/bb8R5

[MartGonzo]

I posted this on the tweakscale thread and thought I should post it here too in the hopes of finding what the issue is.

I've been getting an issue with tweakscale and Infernal robotics, not sure what's causing it so I wasn't sure where to post it.

anyway I use infernal robotics parts on satellites and when changing craft and going back later sometimes the parts are moved, sometimes to the center of the craft or sometimes offset completely and floating a meter or so from the satellite.

Has anyone come across this or know of a fix?

[allmhuran]

I posted this on the tweakscale thread and thought I should post it here too in the hopes of finding what the issue is.

I've been getting an issue with tweakscale and Infernal robotics, not sure what's causing it so I wasn't sure where to post it.

anyway I use infernal robotics parts on satellites and when changing craft and going back later sometimes the parts are moved, sometimes to the center of the craft or sometimes offset completely and floating a meter or so from the satellite.

Has anyone come across this or know of a fix?

From what I understand, this is a known issue with KSP in general.

There are three places scale can be set in a part config. The "rescalefactor" (which applies to the entire config), and two places in the "model" reference (for parts that use them).

If the "rescalefactor" is used, then switching away from the craft and back to it sometimes (always?) causes the scaling to be reset.

There is one solution that always works, but you might not like it: build your parts to the scale you need, and then use the welding mod to create new subsections. The welding mod always uses model reference blocks, and always sets the scale inside those blocks, which always works. The problem is that you have to create a bunch of new parts if you use this solution.

For me that's fine, I need to create new welded parts anyway otherwise my craft part count would be ridiculously high. But you might not want to weld parts together. If the only parts you are having issues with are the robotic parts themselves then that might not be so bad since you won't end up with too many redundant, resized parts.

Another possible solution that you may not have considered: parts being transformed by IR servos can clip through each other. This works reliably *as long as the part is always clipped*. If at any point it is not clipped, you may have issues when it tries to intersect again through the movement.

In other words, to make your joints bigger you can actually encase the IR part and its attached parts entirely within, say, a fuel tank. I use exactly this solution on the "knees" of my new Timber Wolf (look at the joints half way down the leg in this video:

) Edited August 24, 2015 by allmhuran

[Tarheel1999]

I posted this on the tweakscale thread and thought I should post it here too in the hopes of finding what the issue is.

I've been getting an issue with tweakscale and Infernal robotics, not sure what's causing it so I wasn't sure where to post it.

anyway I use infernal robotics parts on satellites and when changing craft and going back later sometimes the parts are moved, sometimes to the center of the craft or sometimes offset completely and floating a meter or so from the satellite.

Has anyone come across this or know of a fix?

Docking can cause this or similar.

[MartGonzo]

From what I understand, this is a known issue with KSP in general.

There are three places scale can be set in a part config. The "rescalefactor" (which applies to the entire config), and two places in the "model" reference (for parts that use them).

If the "rescalefactor" is used, then switching away from the craft and back to it sometimes (always?) causes the scaling to be reset.

There is one solution that always works, but you might not like it: build your parts to the scale you need, and then use the welding mod to create new subsections. The welding mod always uses model reference blocks, and always sets the scale inside those blocks, which always works. The problem is that you have to create a bunch of new parts if you use this solution.

For me that's fine, I need to create new welded parts anyway otherwise my craft part count would be ridiculously high. But you might not want to weld parts together. If the only parts you are having issues with are the robotic parts themselves then that might not be so bad since you won't end up with too many redundant, resized parts.

Another possible solution that you may not have considered: parts being transformed by IR servos can clip through each other. This works reliably *as long as the part is always clipped*. If at any point it is not clipped, you may have issues when it tries to intersect again through the movement.

In other words, to make your joints bigger you can actually encase the IR part and its attached parts entirely within, say, a fuel tank. I use exactly this solution on the "knees" of my new Timber Wolf (look at the joints half way down the leg in this video:

)

Thanks for the reply but scaling isn't the issue, the parts aren't being rescaled they are being moved and it only ever happens with IR parts

I think these screenshots should demonstrate what I'm talking about

Javascript is disabled. View full album

In these satellites the scansat equipment is connected to pistons and the high res scansat equipment is connected to hinges, most of the satellites are fine, but on these particular ones the IR parts have been moved inside the craft, they still works as intended but are in the wrong place.

I don't think using the welding mod would help as it's only the IR parts affected and presumable welding them would stop them working and I really don't think it's a stock issue as I've never seen parts move to a different position before.

Docking can cause this or similar.

The satellites haven't docked to anything

[allmhuran]

Thanks for the reply but scaling isn't the issue, the parts aren't being rescaled they are being moved and it only ever happens with IR parts

http://imgur.com/a/aFHxZ

Indeed, but if the parent part is rescaled then the children, which list that part as their attachment reference, may move to new positions based on the new parent scaling. I can't tell if that's what is going on in those pictures but it's worth a look. Or are you not rescaling anything?

[MartGonzo]

Indeed, but if the parent part is rescaled then the children, which list that part as their attachment reference, may move to new positions based on the new parent scaling. I can't tell if that's what is going on in those pictures but it's worth a look. Or are you not rescaling anything?

Ah I see what you mean, although to be honest I'm not sure that would make sense as it doesn't happen with all the satellites I've launched, just a couple, and they're all the same design, and plus, the strut they were attached to was scaled down yet the parts move inside the craft, if it were to do with the original scale wouldn't they move outward?

anyway yes pretty much all the parts of the satellites are rescaled except the root, I will try attaching the IR parts to the probe core and then using offset.

unfortunately it's not an easy problem to replicate as it only happens once in a while so I won't be able to say definitively that that was the cause unless it still happens.

Thanks for the suggestion.

[Spooglecraft]

While testing my Star-Wars-like Astromech Droid I noticed a missing face on the rotatron model. (Seen in picture.)

[allmhuran]

That's the bottom

[Garbonzo]

Hey, I downloaded this and I didn't see any of the white-orange parts. I noticed that they were .90 mods. Are those safe to download or should they be avoided? and if they should be avoided will they be updated?

[allmhuran]

The rework parts have caused no additional problems for me in 1.04