[WIP] MSI's Infernal Robotics - Model Rework (v02 Pre-Release Beta)

[BudgetHedgehog]

No no, carry on! I might not even be able to do it, you go ahead, mate

[ZodiusInfuser]

I welcome as many people helping out with logo creation as possible . I imagine there'll be a few rounds of iteration.

In the meantime...

(yes I know I've got wheels to do too )

I'd welcome any feedback on the design at this stage, as it's still in-progress, but I felt it was too awesome not to show.

[*Aqua*]

It's already great!

I think it'll look good if the trailer part and the end part cut into each other. Something like this.

Or the other way around: Add something to the trailer and cut out something at the ends.

[CaptainKipard]

It's already great!

I think it'll look good if the trailer part and the end part cut into each other. Something like this.

http://i.imgur.com/r2USmDJ.png

Or the other way around: Add something to the trailer and cut out something at the ends.

That to me looks like it would slightly compromise structural integrity IRL and therefore break immersion.

[Darren9]

Personally I prefer long thin ones to short fat ones It could probably be a bit longer for it's width? Looks great already though.

[*Aqua*]

@Cpt. Kipard

In rl you would have cushions at the end so it couldn't ram into the end with full force.

[Horus]

In the meantime...

https://www.cubbyusercontent.com/pl/IR_GantryPreview.png/_d371567bdd094d8282bf486aa0877bdd

(yes I know I've got wheels to do too )

Well, now I'm 100% sure that I do not need Legacy IR parts when all of these are done

[ZodiusInfuser]

That to me looks like it would slightly compromise structural integrity IRL and therefore break immersion.

I'll play about with the design and see if something can be done, since I've realised it's slightly too long for a particular application I have in mind anyway .

Personally I prefer long thin ones to short fat ones It could probably be a bit longer for it's width? Looks great already though.

I'm planning to do two different lengths of gantry for the 3 different versions I make. The one in the picture is the "Slidertron - Basic" in 2.5m. There'll be a 5m version of it, as well as a side mounted and a reverse mounted. This is in addition to the 3 tweakscale sizes.

Edit: Actually I might as well ask now, what lengths would people like to see? Bear in mind that this isn't intended for large scale shipyards, since it doesn't fit the stock circular profile.

Edited January 31, 2015 by ZodiusInfuser

[riocrokite]

hey Zodius, let me know what kind of logo symbol do you have in mind and I'll try to prepare something. The more you describe / post sample images the better.

IMHO gantry looks nice and elegant in its current version.

[ZodiusInfuser]

hey Zodius, let me know what kind of logo symbol do you have in mind and I'll try to prepare something. The more you describe / post sample images the better.

You can find details of what I'd like over in this thread: http://forum.kerbalspaceprogram.com/threads/108596-Logo-for-Infernal-Robotics-Model-Rework

IMHO gantry looks nice and elegant in its current version.

I tried adding the feature Aqua suggested, but I couldn't get it looking nice with the shrinkage I had to do to get the gantry to fit in the thing below...

That's the old 1.25m diameter ship truss I modelled ages ago btw

Now to figure out how to do the other gantry types... :|

[CaptainKipard]

On a satisfaction scale how snugly does that fit?

[ZodiusInfuser]

On a satisfaction scale how snugly does that fit?

Probably a 4/5

I modelled the gantry to be as wide as a foldatron, then I remembered that this truss was designed to fit foldatrons. I just forgot to check the length of the space at the start, hence the reason to shrink it. The corners are the tightest part, but I can correct that when it comes to redoing the geometry of the truss.

Edited January 31, 2015 by ZodiusInfuser

[techno65535]

Ahh! That truss! I'd almost forgotten about them. Now I can't stop thinking of Star Trek: First Contact.

[riocrokite]

Zodius, your mailbox is full so I post my reply here:

As description says; all measures of the frame are in blender units. For KSP units you need to multiply everything by 1.25. So under the 'wings' there is only 0.725m (blender) for the wheels. FHX extra-wide frame has a little bit more (0.85m blender). I guess the full rotation diameter won't be needed; my guesstimate is that 20-30 degrees left/right would be sufficient.

I imagine the wheels for frame would have to be specific (quite narrow and with very strong springs since they will have to work in different 'mass per wheel' range than usual rover wheels). So don't worry if they don't fit. Just go ahead with your design if you already like it.

[ZodiusInfuser]

Sorry cleared (a bit) now. Ah you don't override the rescaleFactor like I do. In either case, I imagine a frame like this would be necessary.

[lo-fi]

Something like that would actually be rather cool as a (new) sort of procedural. As the lateral sections are identical, having a little code to spawn extra segments (I mean as objects within the same part) as commanded by a tweakable slider, then adding end sections would not be difficult. Thoughts?

Work with the KF hitch has given me some interesting new thoughts on the mecanum wheel operation too. Can you come up with a set of equations that will calculate correct wheel rotation for the desired motion of the vehicle, taking into account X and Y position and the wheels handedness within the craft? So if the wheel knows its position and orientation relative to the other wheels and in the craft, you feed it a control, say "move laterally left" or "turn right" and it will figure out what it should be doing? Obviously these would need to be summed, but you get the idea. Creating configurable joints is childs play, so if the maths is there, setting a joint up within the wheel and feeding the motor drive a rotation speed will work rather well. Also, suspension is also possible within the joint!! I also have a novel way to cheat a way around the little "bounce" the one dimensional colliders currently give.

There's always the single collider method too - I'm not sure yet which one will yield better results, but I have a suspicion the maths I'm after probably exists somewhere already.

Edited February 2, 2015 by lo-fi

[riocrokite]

Sorry cleared (a bit) now. Ah you don't override the rescaleFactor like I do. In either case, I imagine a frame like this would be necessary.

http://www.etftrucks.eu/images/ETF383web.jpg/

Zodius, do you plan your wheels to have top mount or forward mount?

To all: What would be main benefits of this type of frame?

Edited February 2, 2015 by riocrokite

[ZodiusInfuser]

Something like that would actually be rather cool as a (new) sort of procedural. As the lateral sections are identical, having a little code to spawn extra segments (I mean as objects within the same part) as commanded by a tweakable slider, then adding end sections would not be difficult. Thoughts?

That would be brilliant if it could be done (and if it could instance new attachment nodes). Such tech would be great for turning my truss set into procedural parts, since they have repeating elements that will need to be instanced.

Work with the KF hitch has given me some interesting new thoughts on the mecanum wheel operation too. Can you come up with a set of equations that will calculate correct wheel rotation for the desired motion of the vehicle, taking into account X and Y position and the wheels handedness within the craft? So if the wheel knows its position and orientation relative to the other wheels and in the craft, you feed it a control, say "move laterally left" or "turn right" and it will figure out what it should be doing? Obviously these would need to be summed, but you get the idea. Creating configurable joints is childs play, so if the maths is there, setting a joint up within the wheel and feeding the motor drive a rotation speed will work rather well. Also, suspension is also possible within the joint!! I also have a novel way to cheat a way around the little "bounce" the one dimensional colliders currently give.

I probably could come up with the rules for deriving that, but be warned that a lot of the math overlaps with crab steering and Ackermann correction. Essentially you require the same local velocity vector that all my other wheels would require, just the final interpretation of it changes from "turn wheel to face it" to "apply rotation proportional to the vector along our wheel axis".

I'm still not sure that the multiple collider route will return the best result based on my testing, although maybe your proposed cheat could get around that problem.

Zodius, do you plan your wheels to have top mount or forward mount?

The wheel will mount from the top, so that they have 360 degree uninterrupted rotation capability.

Edited February 2, 2015 by ZodiusInfuser

[lo-fi]

I think that kind of procedural is worth pursuing. As you say, there are lots of cool things that could be produced like that. Riocrokite's frames could be another good example, actually. The other advantage it has would be the lack of wobble between the sections as there would not be any joints involved. The nodes are probably not too difficult, it will just take a little thought and code diving some of the JSI switchable node modules. The biggest sticking point is probably what to do when a frame has been created and parts have already been added. There would have to be some way to lock it down once parts are attached to it, or crazy stuff might happen.

If you can come up with some equations, I'll have a go with both methods for the wheels. It's interesting playing around either way. I'm running behind on the KAS fork with grab animations, btw. I know where I want to insert some code, but have yet to create the animation module. I want to do the same thing for my trailer hitch, so I'm going to write it for that so I'm using my own, much simpler code to develop it against, then I'll be able to paste it into KAS with some confidence. Time is my enemy, but I will get there!

Such cool stuff going on at the moment!

[CaptainKipard]

Something like that would actually be rather cool as a (new) sort of procedural. As the lateral sections are identical, having a little code to spawn extra segments (I mean as objects within the same part) as commanded by a tweakable slider, then adding end sections would not be difficult. Thoughts?

I actually suggested this to the person who made stretchy tanks. I'd be so happy if this happened, I'd love to have tanks use it as well. Please work with whoever maintains that mod on this. Don't make it a separate thing. I think it's called "Procedural Parts" now.

Edited February 2, 2015 by Cpt. Kipard

[lo-fi]

Hmmm, this seems a promising avenue to explore...

[ZodiusInfuser]

I Haz logo! Thanks again to riocrokite for creating it

[AntiMatter001]

Uhm... hey... sorryy but i would like to ask a favour...

does anyone know how to scale this down one more time? (extra small) i'm using the RollCage mod and i've made a rover that lasts quite a while on solar power but when i tried adding a claw(not "The Claw" from vanilla) onto the back i relised the parts were WAY too big... and the roll cage scaled up just looks stupid for kerbals...

would love help. if no one can help then that's okay i'll just go back to building slightly less elegant robotic cars

EDIT:on a different note the untility grabber doesn't have a tweak scale value neither does the landing foot or the grasper.

if someone could tell me how to do this i'd be very thankful

EDIT2: just saw the "these parts are still in beta" sentence in the thread... goddamnit... ok i'd like to know why the untility grabber/magnet doesn't have a tweak scale thingy but other than that i guess i'll have to wait :/

Edited February 3, 2015 by AntiMatter001

[Master Tao]

does anyone know how to scale this down one more time? (extra small)

You can do that by editing the appropriate SCALETYPE section of KSP/GameData/MagicSmokeIndustries/Parts/Rework/Rework_TweakScale.cfg. You'll also need to edit the TWEAKSCALEEXPONENTS sections of each part config. The TweakScale OP has the most specific information how to do that, but you could try just scaling down the existing data to add new smaller entries.

i'd like to know why the untility grabber/magnet doesn't have a tweak scale thingy but other than that i guess i'll have to wait :/

This was intentional. ZodiusInfuser wanted to have different utility parts for each size (small: probe, medium: grabber, large: landing foot).

[ZodiusInfuser]

This was intentional. ZodiusInfuser wanted to have different utility parts for each size (small: probe, medium: grabber, large: landing foot).

As Tao says, this is a choice I made as I wanted each size to have different tools that make sense for it. For instance a surface sampler doesn't make sense as a 0.5m part. Now if you can make a case for some of the parts to have different sizes (and by case I mean show me what tweakscaling the part lets you do that you couldn't before) then I may consider adding the odd one. Or if its just the one situation, make a differently styled part for that purpose.

This is also a good point to forewarn people that the currently planned wheel collection will NOT have tweakscale, for a similar reason to the above.

Edited February 4, 2015 by ZodiusInfuser