Orbital Construction Facility

[Tyko]

The benefits of the orbital workshop are the precise placement, access to tools like CoM, CoT and RCS (if using the mod). Also though, the ability to disassemble is really valuable. There's no current way to swap out a module that's not on the "outside" of your craft. Here are a couple of examples of things you could do in a workshop that would be really tough if not impossible with KIS or some sort of enhanced docking:

1. You get nuclear tech and want to upgrade your ship from chemical to nuke.
   1. change the tanks storage from LFO to LF only
   2. pull the old chem engines and attach the nukes.
2. You're probe tech increases and you want to upgrade a probe with a more advanced probe core - these are often buried in the stack and it would be a real pain to remove and add back each piece above it in the stack.

[SinBad]

Ah, you want to be able to retrofit, not just build. i get it now. There is always the option to dock, shut the door, load the mk2 craft file to the shipyard, then it converts your ship to generics, and builds the new one out of those generics adding or subtracting the difference to the storage. You would still be doing the modification in the VAB, but you wouldnt have to launch the mk2 and deorbit the mk1.

[Tyko]

14 minutes ago, SinBad said:

What about a close integration of kis/kas and infernal robotics? All we really need is a kas capable robotic manipulator. Have it remove the limitation of engineers per ton as well. Then we could all build gantry mounted 7 axis manipulator arms.
 and build our rockets out of kis containers.

On the subject of packability, i remember something from the 90's about nasa testing aluminium extruders in orbit. Most of the large pieces folks are worried about are the most packable. Jumbo tank is just a bunch of extrudec struts welded togeter with sheeting welded over the top. You can squeeze quit a few square meters of 2mm thick sheet metal into a cubic meter. Then you just extrude your pipe work, get some cables off the reel and a few sensors an selonoid valves out the bottom drawer and get to work fitting it. Im not saying that we as the player should have to do all of this manually, just that when you put a jumbo into one of the huge inline kis crates, what you are putting in is the materials required to fabricate it (it weighs the same, and takes up about the right volume for disasembled struts and sheet metal). That fabrication happens automatically when you take it out again, otherwise it would be a pita to do anything.

EPL works the same way, just allows you to keep a stock of generic materials. So instead of having a stack of sheet metal that can only be used to make a jumbo tank, you have a lump of metal that could be extruded into struts, or sheets, or pipes, or cables. To avoid pita, its just callled rocket parts.

some really cool thoughts here. The VAB still offers all the the placement tools and if you just create a VAB in space, most of the coding is already in place without writing a whole bunch of new systems.

the overly simplistic view but:

• you'd need a way for an existing craft to be imported back into the VAB environment - once in there, the VAB already has everything else
• a part, or series of parts that you have to orbit to create the "import node" that provides the ability to import the craft into the orbital VAB
  • this would be your "spacedock" and the capacity of the spacedock in parts / tonnage (whichever is easier) would be dictated by how big your spacedock was.
  • Maybe each spacedock component accommodates 10 tons/20 parts (or whatever, I'm just throwing out numbers).
  • If you want to work on a 100 ton ship, you need to launch and assemble 10 of those spacedock parts.
  • Spacedock parts would require power, crew etc. - these are all mechanics built into the game already

I'm proposing this because it seems like a fairly straightforward thing to create if you can just reuse all the VAB functionality. And, why wouldn't it be cool to bring a ship in and upgrade it after long mission? This happens IRL to naval vessels and I would expect that we'd do it IRL in space in the future.

Edited July 28, 2016 by tjt

[Nansuchao]

What if you can build ships from Ore? Maybe, as a end tech tree gift, you'll gain the ability to deliver Ore to a station and with a huge amount of Ore and electric charge you can build your ships, slowly.

[Tyko]

1 hour ago, Nansuchao said:

What if you can build ships from Ore? Maybe, as a end tech tree gift, you'll gain the ability to deliver Ore to a station and with a huge amount of Ore and electric charge you can build your ships, slowly.

I was thinking about this earlier actually. You can already turn ore into parts with Extraplanetary Launch Pads. The goal we are working on here is the "assembly" after launch part. If you had a mod that let you build parts from ore, you could still launch those parts up to the spacedock

[Van Disaster]

On 07/28/2016 at 0:35 PM, pandaman said:

You could yes, and that's a nice simple and elegant solution, but the VAB functions also allow precise rotation and orientation which is 'a wee bit fiddly' with just docking ports.

Well, I'd not suggest the ports immediately vanish on docking - they're there as intermediaries and in that position *they* could let you rotate and translate the new part. Once you're happy, finalize the connection and the vessel tree is rebuilt without the ports or the subvessel stuff that comes with docking.

I'm actually wondering if you could extend attach nodes to do this rather than docking ports.

Edited July 29, 2016 by Van Disaster

[tater]

I have no personal desire to retrofit, as honestly I don't think that should be a thing. I'd be fine with assembling sub-assemblies, and adding parts small enough to be packed in a KIS container (solar panels, experiments, and other stuff to get added to the outside). I'd personally never add/change engines.

[tater]

Here's a workshop concept that has other uses. Allow lassoing parts in the VAB (an entirely normal UI thing that should ideally e a thing, anyway). You then have a group of parts selected you can move around. If a craft is saved already, allow the lassoed parts to be saved as craft-specific sub assemblies (a new tab).

So you create a huge "Jool Explorer" craft that is too big to launch at once on a rocket that actually looks like a rocket (sorry, my bias is showing here, I never make anything that doesn't pretty much look like a real rocket). You could add a bunch of 2.5m clamp-o-tons to it, and try to launch in sections, but for the sake of argument let's say it has laterally mounted stuff that makes it wide, and that docking it precisely would be hard as a result).

Instead, you lasso parts which get saved as Jool Explorer:sub1of7, Jool Explorer:sub2of7, etc. You launch them, and have them grabbed by the "orbital workshop" (a large truss section with grabber arms along it capable of grabbing X number of sub assemblies. The Orbital Workshop would see that they are craft "Jool Explorer" and sub parts 1, 2, 3, etc of 7 total. Once all are delivered, it could load the completed craft and "launch" it (which would simultaneously delete all the grabbed subassemblies).

 

[pandaman]

@tater. I like that idea.  

It may actually be an easier implementation than my 'VAB functions in orbit' suggestion.

[Codraroll]

I'm a huge advocate for orbital and extraplanetary base building. Implemented correctly, I believe it would solve many of the challenges and problems the game faces.

First of all, it removes the tedium of getting off Kerbin. After a couple hundred launches, getting out of the Kerbal gravity well is pure routine work. Even with stuff like MechJeb, it is a time-consuming, repetitive task that offers little challenge once you're familiar with the basics of the game.

Second, it gives players a goal to strive for. Your off-Kerbin construction yard would be something you earn, not something you get at the start of the game. Getting a base in Kerbin orbit - or even on the surface of another body - would require an ungodly amount of Science and Funds, and the launch of several very heavy modules. There might even be Rep requirements to even getting the projects approved. It would not be a shortcut on the learning curve, merely a reward for your hard work and experience. Extraplanetary bases could be fixed to a few pre-programmed locations on each body, requiring players to survey the planet and send Kerbals there to plant flags before construction could begin.

Third, it makes longer missions more convenient. Perhaps more easy too, but that doesn't have to be a bad thing. There's a vast skill/time investment gap between working in Kerbin's SOI versus missions to other planets, and a lot more infuriating when something goes wrong. Getting stuck on the Mun because you lack the Delta-V to get off its surface is very frustrating. Getting stuck on Duna for the same reason is much, much worse, since you spent a lot more time getting there, and having to re-do the mission from scratch because of a small design mistake is very annoying considering the time it took before you could discover that mistake (for a seasoned player, it's a small annoyance. For a newbie, it's a "close-the-game-and-don't-pick-it-up-again-for-months"-inducing frustration). If you could launch from Kerbin orbit or even Minmus, it'd be easier to make purpose-built craft for interplanetary voyages, and it would take a lot less time to do off-Kerbin missions.

Fourth, as a result of point 3, off-Kerbin bases act as "stepping stones", encouraging the exploration of the outer planets of the Kerbin system. Going to Jool is currently a daunting task, requiring hours of play and several in-game years. As a result, only a marginal fraction of KSP players ever go there (there was a thread on this a while back). Using a construction/launch facility on Duna makes the Jool trip a lot less demanding. And once bases are established on the Joolian moons, Eeloo is within feasible range, even for a casual player. At the moment, Eeloo is so remote that there's no real need for outer planets - very few players even go halfway there at all, and missions are dastardly hard and time-consuming.

Fifth, it gives Funds a late-game purpose. After a few dozen successful missions, you have enough money to make the resource practically worthless. Once all the KSC facilities are upgraded, and you keep getting rewarded for your exploration, you have to try hard to get the Funds counter down towards 0 again. But merely constructing a basic off-planet facility would cost an arm and a leg. Now, upgrading your VAB on Ike would make the Funds you spent on the KSC seem like pocket change, and the sum you have to lay down to build a facility on Vall... As you unlock the entire tech tree, you go from Science-seeking missions to Funds-seeking ones. Extra con-yards would give all those Funds somewhere to be spent.

Sixth, it gives purpose to all the other space programs that appear to be flying in Kerbin's skies. A host of different actors manage to put one-man capsules in orbit around the Kerbal bodies, or even on the planets. If Rockomax can put ships on the Mun, they could probably fly there with parts for my rocket building facility too. They'd ask for handsome payments, though. While a solar panel costs very little in the KSC VAB, it'd be a lot more expensive in orbit, with prices increasing exponentially the further you go from Kerbin. Building a 10-ton vessel A on Duna would cost twice as much as vessel B on Kerbin, with the capability of delivering vessel A in its entirety to Duna. Or in more Earthly terms, building the Apollo lander on the Moon would cost more than building and launching a Saturn V on Earth. I suppose your off-planet bases could have buildings or facilities which reduced these costs, though.

Seventh, the system wouldn't be forced on anybody. It's a single-player game, who would you be cheating? It wouldn't remove the option of launching all your missions from the KSC either. Play the game your way, there's no competition and nobody judging you.

 

I would even be satisfied with Squad if they just added support for such a system in their game. Namely, the ability to have multiple KSCs, even on locations off Kerbin. If I understand correctly, this is currently not possible, although you can change which planet the game considers to be Kerbin. Just lay down the groundwork, and modders could take care of the rest. Properly integrated extraplanetary bases and eventual colonization would really add longevity and purpose to the game, which at the moment turns into a bit of a sandbox once the tech tree is unlocked.

[Van Disaster]

Just now, Codraroll said:

I would even be satisfied with Squad if they just added support for such a system in their game. Namely, the ability to have multiple KSCs, even on locations off Kerbin. If I understand correctly, this is currently not possible, although you can change which planet the game considers to be Kerbin. Just lay down the groundwork, and modders could take care of the rest. Properly integrated extraplanetary bases and eventual colonization would really add longevity and purpose to the game, which at the moment turns into a bit of a sandbox once the tech tree is unlocked.

You're somewhat mistaken - as has been brought up repeatedly in the thread, investigate Extraplanetary Launchpads. I'm not sure about the current state of PQSCity spawnpoints ( the system KerbalKonstructs uses ) but given EPL can spawn and destroy vessels anywhere including in orbit, I've no reason to believe KK wouldn't be able to *somehow* if it had multiplanet support.

[DStaal]

On 7/29/2016 at 11:17 AM, tater said:

I have no personal desire to retrofit, as honestly I don't think that should be a thing. I'd be fine with assembling sub-assemblies, and adding parts small enough to be packed in a KIS container (solar panels, experiments, and other stuff to get added to the outside). I'd personally never add/change engines.

You need to take a look at a mod that's been overlooked in this discussion: OSE Workshop.  It allows you to '3d print' small parts directly into KIS inventory.  (From a similar resource chain to EL+UKS.  Actually, RoverDude switched to it's resource chain when it came out, I believe.)  I'm not sure if it can do sub-assemblies, but I think it might be able to.  Upper limit on size is generally dependent on your KIS storage, but partly dependent on your workshop.  You then have to use KIS to deploy them.

On a general note in this discussion: EL can de-spawn ships too.  It calls it recycling, and it despawns one part at a time.  Conceptually, the 'retrofit a ship' idea is basically the same as 'recycle a whole ship and rebuild a nearly identical one', though a mod that specifically did the former would be an interesting idea - and something I can see coming earlier in the tech tree.  But it would probably be a good idea to co-ordinate efforts with EL, as they have much of the code you need and it'd be nice if the two played nice together.

I don't necessarily think you need to actually change the VAB/SPH as you do that - EL treats them as basically design spaces and testing grounds; everything is designed and tested there, no matter where the final build is intended to be.

[Tyko]

10 hours ago, DStaal said:

You need to take a look at a mod that's been overlooked in this discussion: OSE Workshop.  It allows you to '3d print' small parts directly into KIS inventory.  (From a similar resource chain to EL+UKS.  Actually, RoverDude switched to it's resource chain when it came out, I believe.)  I'm not sure if it can do sub-assemblies, but I think it might be able to.  Upper limit on size is generally dependent on your KIS storage, but partly dependent on your workshop.  You then have to use KIS to deploy them.

On a general note in this discussion: EL can de-spawn ships too.  It calls it recycling, and it despawns one part at a time.  Conceptually, the 'retrofit a ship' idea is basically the same as 'recycle a whole ship and rebuild a nearly identical one', though a mod that specifically did the former would be an interesting idea - and something I can see coming earlier in the tech tree.  But it would probably be a good idea to co-ordinate efforts with EL, as they have much of the code you need and it'd be nice if the two played nice together.

I don't necessarily think you need to actually change the VAB/SPH as you do that - EL treats them as basically design spaces and testing grounds; everything is designed and tested there, no matter where the final build is intended to be.

3D printing a fission reactor feels too much like Star Trek Replicator technology for my taste. The whole "retrofit" idea is more about taking an existing ship and replacing/adding/removing parts.

Here's a great example. I have a ship that was built to conduct landing and science missions over Minmus and the Mun. It included a big, heavy science lab at it's core because that was the purpose of the ship. It's a year later and I'm done with lunar science, but would like to extend the life of the ship to use it as a mothership for base building. Ideally I'd cut the lab out because continually carrying that extra mass around is wasteful. In real life ships are modified like this all the time, but there's no way to do it in KSP.

(P.S. I know that having built it initially with docking clamps could have been an option and I did use docking clamps for some parts, but you can't build a ship with docking clamps between each part, the part counts go through the roof and you ships get all wobbly.)

[DStaal]

I'd argue that if we ever do build a working fusion reactor, it's likely to be using some 3d-printing type of technology.  (Though I don't think I know of any fusion reactors that are small enough that OSE Workshop can do them.)

For your ship - you should also take a look at WildBlueIndustries's collection of mods.  If you'd built that ship based on a MOLE instead of a standard science lab, you could convert it over to some other module now.  (At an optional resource cost.)

[Van Disaster]

I'd have built that ship in modules, connected with docking ports so I can replace propulsion sections / pull the science sections off etc - NOT connecting every piece! just discrete units. I do that all the time, fuel+engine+rcs tanks+probecore+electrical behind a docking port, so I can both use the propulsion section as a tug & swap it out for a more efficient one later. You can do all that in stock. although using KJR/eva strut mods really help. If you look at how we build real spacecraft of any size these days, they're all built of sections connected with docking ports too; Mir, ISS, any proposals for anything of significant size. KSP *is* capable of keeping multi-module craft in one piece provided you obey the same rules you would building a rocket.

[kdkasad]

On 7/23/2016 at 1:40 PM, Tyko said:

A clamp-o-tron allows you to move modules around, but I don't see that a orbital construction. I'd love to have an orbital VAB where I can pull ships in and do refits on them.

Speaking of which - is there a way to change Action Groups outside of VAB?

There is a mod called Action Group Manager that lets you do that.

[Youen]

I know this thread is old, but if anyone reads this interesting discussion like I did after landing it from your favorite search engine, I'd like to add an existing solution that wasn't mentioned : the Konstruction mod allows to weld* docking ports. The process is then:

1. design your ship/base/space station in VAB as you like
2. cut it in manageable blocks for launch, and add docking ports and RCS/probe cores as needed
3. Launch the blocks
4. Dock the blocks together in orbit
5. Separate unneeded RCS/probe cores (KAS can help here)
6. Then you can right-click each docking ports and select the option to collapse them. This will remove both docking ports, and attach the corresponding parts together, exactly as if it was done in the VAB.

Beware, this is an irreversible process, no undocking after welding (the docking ports are completely removed from the ship)

This has the advantage of looking nice, and reduce the part count (and reduce wobble if you don't already use a mod for that).

Now about practical aspects if someone is interested:

• Konstruction can be installed with CKAN or from here (the thread says KSP 1.3 but it works fine in 1.6)
• (optional) I'm trying to avoid getting too many parts in my KSP install, so I've just kept Konstruction.dll (and the dependency USITools.dll), and added ModuleWeldablePort to stock docking ports using a Module Manager config file. I've deleted all other parts from the Konstruction mod. Maybe I could release that as a stand-alone mod.
• If you need to precisely adjust roll when docking, while there are options for that in Konstruction, I've found it doesn't work so well for precision work. So I do it manually, right before docking, which feels more "role play" actually. I use the Docking Port Alignment Indicator mod, which, among other things, adds a roll angle (visual and numerical) indicator, making it easy to dock with sub-degree accuracy. When done right, you don't feel the "magnetic" attraction, your block just clicks into place seamlessly. Beware, as of Konstruction version 1.2.0.0, there is a bug with the "Compress Parts (Rotate)" option which is supposed to adjust roll when welding, but actually increases it. But if that bug gets fixed, it would probably be the best way to go.

I still have to put all this into practice to get my Mun base assembled and landed, but from the few tests I've made it should work well.

Happy assembling !

 

* the term "weld" here just means the parts get attached together, exactly like when you do so in the VAB, it won't merge parts into a single one.

Edited September 16, 2019 by Youen