thereaverofdarkness2

It has to do with the view angle. The edge is always distorted relative to the center, it's just that with a low enough view angle you can't see it anymore. KSP rendering is not escaping this problem--because that's impossible--rather it is circumventing it in a way that drives me up the wall and apparently most people feel just fine with.

The wheels explode because they read their contact spin velocity as being equal to the craft's surface velocity relative to the ground underneath it. I'm almost certain it would be a simple matter to program in a better method for determining spin speed, but more importantly there's little problem with simply allowing them to turn at infinite RPM like landing gear, and letting their impact velocity determine whether or not they explode. I'd like to see all parts capable of reaching a broken state before they explode. It could be neat if part of your rocket breaks and becomes inoperable but it's still attached and so you still have to carry the mass. But stabilize it and send your engineer out on a spacewalk and you can fix it. The landing gear is unrelated and I don't have a clue why it's exploding. I've not had problems with landing gear exploding, but you might check to make sure no other parts are clipping through the landing gear.

What you're referring to as a "fisheye" lens is a wide angle view, and there is no way to avoid what you call distortion. Every flat rendition of a 3D image is distorted and the "fisheye" wide angle lens is a close representation of what we actually see when looking around. Our brains interpret this flat image differently from how they interpret the distortion we see in our peripheral vision. You may feel like the scenery in KSP matches real life, but it does not, and this is easily seen by trying to compare it with real life. If you try to compare it by memory, it seems right (to you guys, apparently). But make a direct comparison and you will see that it is way off. There's no way to truly match the screen graphics to what we see with our eyes, but I've never seen any other 3D game with the rendering style used in KSP. Perhaps it is this lack of barrel distortion as cantab said it. Our brains gloss over barrel distortion in our periphery, so at first it may seem right to eliminate it from the game. But putting barrel distortion into the game produces a more accurate image, it matches what real-world cameras will show, and you can get used to it over time. Most importantly, virtually every true 3D rendering game ever has some sort of barrel distortion...I'll name a handful of examples: Minecraft, EVE: Online, the Quake series, everything built with the Unreal engine, Half Life 1/2 and Portal 1/2, the list goes on and on and on. But examples of games without this distortion are rare and seem to be relegated to false 3D renders such as Wolfenstein 3D, the original Doom, or Duke Nukem 3D. KSP is the only true 3D game I can think of that renders this way.

I don't like it at all, but I seem to be in the minority.

No, it has nothing to do with it. Let me show you an image from 30km over Kerbin, and let's compare to 300km over Earth: Compare that to the Challenger spacewalk image above. See anything different? They are at comparable altitudes relative to the size of their planets.

That has nothing to do with it. A view from 80km altitude over Kerbin should have the same horizon as a view from 800km altitude over Earth--it's not much different than 300km (ISS altitude). It's not till around 2000km over Earth that it starts to look less like a horizon and more like a ball. Image from LEO using normal view angle, altitude unknown, likely 200-400km Image from LKO using default view angle, altitude 80km

I want KSP to look more like the view from our eyes. It doesn't look correct at all. It looks very, very wrong. Another good way to show this is to get into low Kerbin orbit-80km or so. From there you can see a VERY distinct curvature to Kerbin. No matter which direction you point the camera, the horizon curves steeply away at the edges, even with a very narrow view angle. In a realistic view at that altitude the horizon would look nearly flat, almost like you were still on the ground.

The view doesn't look right from any angle. I used wide field to make the discrepancy more obvious, but it's quite present at the default camera setting.

This has been bugging me for a while, but KSP never looks anything like actual camera views. Today I discovered how to alter the view angle of the camera, and after playing around with it for a few minutes, I believe I am certain that this game simply does not render an accurate 3D model of the terrain and planets. The easiest discrepancy I can think of to point out is that the horizon maintains the same curvature on your screen relative to your altitude, regardless of the direction the camera is facing. In a real-world camera, the horizon always looks flat as you look directly at it; it curves downward at the sides as you look up, and it curves upward as you look down. This is most easily seen in a wide-angle "fish-eye" lens, but the effect happens at any view angle. KSP does not emulate this at all. It seems to be building a cylindrical image around the camera, rather than a spherical image. Looking up and down simply moves the world down or up, but doesn't actually distort it the way it should be distorted when you look down or up--the result is that it appears distorted because it's not distorting when it should be. EDIT: It has been explained to me that KSP and all other 3D games generate all straight lines as straight lines on screen. This explains why the horizon curvature doesn't change with your view angle. But the rendering in KSP is different from the rendering in other games, in different ways. It is these differences that are bothering me, that nobody else seems to notice. They show up when you compare KSP to other 3D games. Please discuss this topic, and ignore straight lines vs. curved lines. ALSO: Several people have mentioned that they don't want to have a fish-eye effect in the game. I was never suggesting at any point that wide-angle distortion is something the game should have at narrow view angles, but it already happens at wide angle view. Please don't discuss whether or not you want distortion, because it's not even related to the topic. The distortion in the images below is due to them being wide-angle shots and cannot be prevented without severe damage to the image quality. If the KSP images don't look distorted to you--well they are distorted, but it's just further evidence of my point. In other 3D games, a wide angle makes everything look highly distorted. How to change view angle: In KSP you can adjust your view angle by holding alt and rolling the mouse wheel. In Minecraft the setting is in the basic menu in the upper left. If it says "Normal" or "Quake Pro" that means either 70º or 110º, respectively, but you can use the slider to change it. In all Quake games there is a slider in the menu to change the field of view, and I believe it lets you go all the way to 180º but it's not marked by degrees. I don't really know how to explain this, so let me just show you some real-world images compared to similar images taken from KSP: Wide field view pointed below horizon Wide field view pointed below horizon in KSP Wide field view pointed at horizon Wide field view pointed at horizon in KSP Wide field view pointed above horizon Wide field view pointed above horizon in KSP

I don't think of this often because I am so familiar with Kerbin's terrain that I can immediately spot where the KSC is, provided it's on the day side. This is annoying when it's on the night side because there are no night lights to reveal the coastline (I need to get a city lights mod). But when I was newer and as-yet unfamiliar with Kerbin's terrain, it was very difficult to find the KSC. I think it would be great to have a KSC pin on the map and, provided there is a way to turn it off, I can't think of any downsides.

The main planets and moons should already be known, but it would be neat to have KBO discovery. There should be a rare chance of discovering a giant KBO, and it would always be different for each game, so you could have a unique planet to show off to the rest of the KSP community!

You can disable steering to disallow the wheels to swivel side to side. Best to disable steering in the VAB so that you know the wheels are locked straight and not canted slightly.

Yes, please!

Nothing I have suggested here makes science collection more difficult than it already is. The base return values are the same as what you already get, only the max amounts are increased. It would be neat to have some of this made easier or harder based on your difficulty setting, but ultimately there's a lot of power to the player to define their game by choosing how much effort they want to put into gathering science. The Science, Jr. hasn't been nerfed. You can get just as much science as before without using a Science, Sr.--it's up to you to bring that monstrosity if you want it. I picture the Mobile Processing Unit (MPU) (the small one) being similar in size to the Science, Sr., about the size of a 2.5m monopropellant tank and probably having around 0.5 to 1.0 ton mass. The currently existing Mobile Processing Laboratory (MPL) is 2.5m wide and very tall, with a mass of 3.5 tons. The Mobile Research Laboratory (MRL) would be 3.75m wide and have a height similar to the MPL, probably giving it a mass somewhere around 8 tons. The names are a bit confusing but very much tentative... I'm not sure how big the Mystery Goo Habitat should be, but I was imagining a short 1.25m dome module you could attach to the top of a stack, when it opens it reveals a clear dome window through which you can see the goo habitat innards, with the goo moving about weirdly in some slow animation. Maybe similar size and mass to the Science, Jr. Maxing out the gravity reading through setting up seismic stations would be very tedious, but there's no need to do this. With the advanced instrument, you collect 100% data on recovery. Even with just the basic module, you still get 80%. There's only a small gain from monitoring over an extended period of time. However I was thinking that since the orbiting craft would pass over multiple gravity biomes, it should only collect data while traveling over the right biome. So, then, you could set up an orbiting probe with multiple gravity readers collecting data on all biome types as it passes over each one. To match this for the ground, perhaps a seismometer could very slowly catch readings for biomes it's not in, with the collection rate sped up for every other seismometer actively reading while on the same world but in another biome. So if you land 3-4 craft in different places with seismometers, you could just sit there and gradually soak up all biomes worth of seismic activity. " So when you transmit data and a Scientist is onboard you'll get the transmit bonuses, but only if you transmit that data right? " The scientist attaches the bonus to the data unit when it is gathered. You simply need to have the scientist on-board the vessel when the experiment is run. After that, non-scientists can transfer the data by hand without diminishing it's value. You need to also have scientists inside the processing unit to get the maximum transmit value. As for realism regarding parts and engines: real-world rocket engines have improved a lot over time. We actually improved them mostly by learning the most efficient engine bell shapes for various conditions, but the kerbal designs could also vary in terms of materials they're built with. I have always noticed that despite being much smaller than our rockets, kerbal rockets are much more flimsy and less durable as well as being more dense. I have to assume they are using cheap structural materials and that they could improve the designs with better strength to weight ratios as well as more efficient shapes. But early science could unlock cheap, lousy engines more akin to what NASA produced in the early days. Middle science would unlock high-quality conventional engines. And the most advanced science would unlock the really cool engines, like the LV-N or the PB-ION. It doesn't make sense to me that some parts like the Docking Port, Sr., the Kerbodyne adapter, the hubmax multipoint connector, or other structural items are unlocked so late in the tech tree when they're really just a bigger piece of metal. There should be early versions, though they might be adversely heavy and not as strong as is ideal for their size, but it shouldn't take till the end of the tech tree just to have all the basic structural elements. It would be great to make Kerbal Attachment System stock.

Image - Example comparison of the proposed modules, to show how I imagine their size. I've had this idea developing in my mind for quite some time now, and I think I'm finally ready to present it in its complete form. This proposal would make significant changes to the way science works, but mostly by adding new things. Science collection in general will feel very familiar to KSP players, but it will be more streamlined and balanced, and make more sense. More total science will be available to collect--I won't show a revamp of the research tree but I'll touch upon some ideas to expand it and make room for the science without adding a ton of new items. All of your data collection science will revolve around a base set of experiment types: Crew Report EVA Report Surface Sample Goo Study Materials Study Thermometer Reading Barometer Reading Seismometer Reading Gravometer Reading Early in the tech tree you will unlock all experiment types in their most basic form, you can easily have them unlocked before you leave Kerbin to go to the Mün or Minmus, but later you will unlock better versions. These may allow you to collect a larger percentage of the maximum science on the first try but they will not increase the maximum amount of science you can get from any given experiment type in any given collection region. This means that if you send out a long mission early in the game, that mission can carry all of the important scientific instruments you'll need along the way. It would be great to take advantage of higher technology, but it won't be mandatory. Just like now, each experiment type/biome combination will have its own maximum cache of science. For instance, there's a limit to how much science you can gather from performing goo studies while landed on the Mün's midlands, which is separate from the science you can gain from thermometer readings while landed in the Mün's midlands, or from goo studies while flying over the Mün's surface. Scientists can increase the science gathered when you collect data, and can slightly increase the maximum science you can gain from an experiment in any given location, but the scientist must be the one performing the experiment to get the bonus. Any science processing unit can be used to increase the transmit value of experiments (except crew/EVA reports which already transmit at 100%) and any kerbal can operate these facilities, but scientists increase the bonus transmit value. No matter what, there will always be a good margin of science you can't get without returning the experiments home, but with a sufficiently large processing array and well-trained scientists, you can become able to transmit the better majority of the science. Data Collection Experiments Here I will list each experiment type along with details and tentative numbers about its maximum yield, transmit value, and some suggestions for later tech upgrades to the collection units: Crew Report: Base maximum value: 6 Transmit value: 100% Recover value: 100% EVA Report: Base maximum value: 8 Transmit value: 100% Recover value: 100% Surface Sample: Base maximum value: 60 Base transmit value: 12.5% Max transmit value: 25% Recover value: 50% A kerbal can carry a 1kg sample of regolith which has a recovery value of 50% of the maximum, and this takes up space and adds mass to the capsule it is stored in. Returning more surface samples from the same biome gives much less science--you have to take home a whole ton of material to get the full 100%, but you can make it to 75% by bringing back just 100kg. Later in the tech tree you will unlock a surface collection unit which can collect the surface material rapidly and can store 100kg of surface sample, and then a storage tank that can hold a full ton of surface sample material and maybe a larger tank that can hold several tons--if you were to gather from multiple biomes. Surface samples provide a way to get extra science from the same places you've already been to. It's generally a lot more productive to explore new places for science, until you start to run out of places to explore. Also, they have the largest gap between maximum transmit value and potential recovery value, so they are your biggest incentive to bring the experiments back home. Goo Study: Base maximum value: 15 Base transmit value: 20% Base processing bonus: +50% Max transmit value: 60% Base recover value: 66.67% The Mystery Goo Containment Unit, like other material-based experiments, has a low base transmit value. It is best recovered. There is a diminishing return on how much of the max you get for recovering it--the first one gives 2/3rds of it, and packing a few extra goo units can squeeze a bit more out if you're willing to run multiple goo experiments per biome. Later in the tech tree you'll unlock the Mystery Goo Habitat, a larger and heavier unit which has more and more lively Mystery Goo inside, and is better for study. The Habitat returns for 100% recover value in one try, and yields half more transmit value, 30% of the max up from 20% but has a much larger transmit size in MITs. Materials Study: Base maximum value: 50 Base transmit value: 15% Base processing bonus: +25% Max transmit value: 45% Base recover value: 50% The Science, Jr. unit offers your first mobile research lab, and these will become a staple for squeezing out as much research as possible from any biome, but they are also the heaviest module to tug around. Initially it recovers just 50% and you can run several to increase the recovery, but it would take an exorbitant amount of Sciences, Jr. to reach 100% due to the diminishing return. However, with just a few you can get significantly higher than 50%. Later in the tech tree you unlock the Science, Sr. which is a much larger 2.5m in-line lab with a base recovery of 80% and a base transmit value of 24%. You can get close to 100% total recovery with just a few of these, but they are a lot bulkier. The Mobile Research Laboratory is a late tech tree unlock, it's a huge 3.75m unit which can perform a materials study at 100% recover or 30% transmit value, it can collect and store some surface material, and it is able to process experiments to increase their transmit value. Finally, it can clean out experiments, including itself. Takes three kerbals to run it at max capacity, but it can run with a minimum of one kerbal. Temperature Reading: Base maximum value: 8 Base transmit value: 50% Base processing bonus: +50% Max transmit value: 100% Base recover value: 100% Pressure Reading: Base maximum value: 12 Base transmit value: 50% Base processing bonus: +50% Max transmit value: 100% Base recover value: 100% The thermometer and barometer are lightweight instruments that can gather a bit of useful data. They have the highest base transmit percent, making them especially useful for missions you aren't returning home, or for when you didn't bring enough instruments for every area you're going past. Later in the tech tree you unlock an infrared scanner and an atmospheric analysis module, these are the advanced versions of these experiment types, and they both have a base transmit value of 75% but a much larger transmit size. I am eliminating the existing atmospheric scan experiment, as I am wrapping it in with the barometric scan, making them the same base experiment. Seismic Reading: Base maximum value: 25 Base transmit value: 40% Base processing bonus: +50% Max transmit value: 100% Base recover value: 80% Gravity Reading: Base maximum value: 25 Base transmit value: 40% Base processing bonus: +50% Max transmit value: 100% Base recover value: 80% The Seismic Accelerometer and Negative Gravioli Detector are very small and light like the thermometer and barometer but they have a much larger data size and transmit at only 40% initially. Later you unlock the larger Seismograph and Negative Gravioli Scan Array, which have a base transmit value of 60% and will always recover for the full 100%. You can alternatively set an accelerometer running at a landed craft/base, or set a gravioli detector running on an orbiting craft/station, and have it gradually increase its recover value over time, to eventually reach 100%. Lab Processing of Data Processing labs can process experiments, increasing their transmit value. They can also clean experiments. These units can be operated by any kerbals, but scientists are able to squeeze extra transmit value out, in addition to their overall science bonus based on their level. The Mobile Processing Laboratory (2.5m) and the Mobile Research Laboratory both have the full processing functionality, but the smaller Mobile Processing Unit (2.5m, much shorter) is only able to process for half the bonus. It is much smaller and lighter, however, and thus can be good to bring along if you can't spare as much mass for your trip. The more kerbals you have working on processing, the faster it'll get done. As long as there is one scientist in the crew, you'll get the full scientist bonus, but the extra based on the scientist's level is taken from the highest level scientist in the crew. The small MPU has room for only one kerbal, but it's still faster than the other units because it doesn't process the data as far. Scientists also make the labs run slightly faster. Kerbal Scientists The full processing value increases the transmit value by 50% over base, but a scientist will get a base 75% increase to transmit value at level 0. At each higher level, a scientist gets +5% processing transmit value, +3% to all science gathered, and +5% faster processing lab operation speed. For example, a level 4 scientist gathering a materials study (Science, Jr.) from the Mün's lowlands: The multiplier for this biome is 4x, so the base 50 on materials study gives us a base max of 200 science for this experiment in this location. The true maximum is 230 with a max-level scientist. The module has a base transmit value of 15% and a base recovery value of 50%. The scientist increases the recovery by 12%, to 56% or 112 science on the first try. It will transmit for only 33.6 science, but we can process it higher. A non-scientist processing the experiment in a full-scale station would get it up to 42 science, but our level 4 scientist can process to +95%--however the materials study experiment type only processes half as high, so our scientist can process it up to +47.5%. The scientist would also add a 12% value to the experiment, but it already has a 12% bonus modifier applied to it since the same scientist gathered it in the first place. The final transmit value is 49.56 science, about a quarter of the base max for recovery. The same scientist could transmit for a higher value on the same experiment if a higher tech module was used to collect the data initially. A change to experience gain: Scientists now gain experience by performing experiments. Each experiment performed in each region gives them experience, and they level up by performing enough experiments in enough different areas. They can reach max level by running experiments only on Kerbin, but there are higher multipliers on more distant worlds. Also, kerbals will gain half of their field experience while still on the mission, applied at KSC midnight. They must return home to gain the other half, but they can level up during long missions. Expanding the Tech Tree I had an idea to both stretch out the tech tree and also give the player more options to begin with: provide more of the important types of modules earlier in the tech tree but have later science improve the designs. Important structural elements should be provided early on, allowing you to construct a spacecraft of any shape you wish, but later research could unlock upgrades to the same shapes, making them lighter and stronger. Even large engines should be available pretty early in the tech tree, but later research would unlock more efficient and powerful engines. Later fuel tanks could have a smaller dry mass fraction, or contain advanced fuel that increases engine specific impulse. Wing surfaces could become lighter and stronger as well as gaining a higher lift rating. Early photovoltaic panels could be stationary, and later panels would rotate to track the sun automatically. Later docking ports need not only be larger, but can also have a stronger grip and be thinner/lighter/more durable. And as I have talked about above, the science experiments are all available early in the tree, but have later variants that improve them--though it is possible to get all the science with the starting experiment units. Changing the tech tree in this way would give the player more freedom early on, and would allow players to plan ahead in their space program rather than forcing them to build crap spacecraft that are quickly made obsolete by new science unlocks. One thing that bothers me about KSP is that I have to explore first before I unlock the parts I need for exploring. I want to build my spacecraft, then send it out on a grand mission, and have that grand mission benefit my space program. It doesn't feel right when I have to unlock the whole tech tree just to prepare for launching my grand mission.

Nice job! I see you took a very direct path down to the surface!

I agree. I feel like the original kerbals were intended to not have any sex, but the insistance of SJW players that they are all male pushed Squad to make a second kerbal sex. They handled it brilliantly, however, and gave the two sexes a very distinct dimorphism without assigning the majority of our cultural gender stereotypes to any of them. The females are rounder and have long hair and that's it. Or add 2 and let the player select one of each job.

Static and procedurally generated are not mutually exclusive. All of the terrain is static and most of it is procedurally generated. Given the number of polygons in the larger craters on the Mün, I'd wager they are, to some extent, procedurally generated. If they are reasonably ovoid or potato-shaped, the gravity modeling inaccuracy won't be that strong and won't hugely impact the quality of the world. Its higher altitudes will have noticeably less gravity than the lower areas and this is what you'd usually get in real life. The main difference between reality and simulation in this case is that in the simulation, the gravity variance by altitude is greater than in real life.

It would be difficult, but it's not impossible. The big hurdle is that each world is an altitude map made from virtual splines sticking straight out from the center of the world, and its center of mass is the same as the center of the splines. They are also procedurally generated using fractal seeds to generate the majority of the terrain including the general style of the terrain. For example Minmus' flats are probably created by having the terrain generator pop the lowlands up to a specific altitude, then another factor smooths out the edge of the flats so there isn't a sharp corner there. Only specific features such as the obelisks or the Kraken corpse were actually constructed entirely by hand. A person could easily make a rough 3D sketch of the lopsided world they want, and then determine the position of its center of mass using volume calculations on the computer. But then making the procedurally generated altitude map cooperate to give you a world with roughly that shape is going to be difficult. I, for one, would love to see a lopsided world added to KSP, so I hope the wonderful artists at Squad decide to show us what they're capable of!

The game doesn't use the part's visual model to determine its aerodynamic properties so it would work fine. Make it look like an aerospike, but make it act like a nosecone.

I see. I didn't realize planes get so much lift on Duna. I'll edit the challenge to say no wings either. Sorry for the mistake, I hope you guys understand. And thanks for the wonderful submissions and demonstrations!

Submissions should get extra points for taking advantage of the Kraken for base-building purposes.

The primary target of this challenge is to see if you can land an ion craft on Duna without using parachutes or wings, but I am also adding Vall in as a secondary target if you want something a bit easier (or harder, depending on how you play KSP). Both should definitely be possible, as I have built an ion probe with a TWR greater than 1 on both worlds. Rules: 1.) It doesn't matter how you get to Duna or Vall orbit. 2.) Start in orbit of Duna or Vall. 3.) Use only ion propulsion and any structural parts you wish. No parachutes, no lifting surfaces, no airbrakes, and no engines other than the Dawn engine. 4.) Using decouplers or other explosives as a source of thrust counts as cheating. 5.) Come to a complete stop with an operational craft. It can be as intact as you want, and it doesn't have to be able to take off again. x.) Mention your version of KSP and any aerodynamic mods you use. It won't affect your score but I want to mention them because it can affect the challenge. If you can't make the landing, I'll score you on impact velocity. The lower your velocity, the higher your score. Successful landings are ranked at the top in order of when the entry was posted here. I won't be doing this challenge myself at current because I cannot play KSP after 1.1, and that makes me very sad. =( Perhaps watching your attempts will cheer me up! If you decide to land an ion craft on another world and you think it was difficult, post it here and I might add a spot for you on the leaderboard! Leaderboard for Duna 1.) astrobond - landed on Duna Defunct Leaderboard These submissions were posted when I allowed wings. They showed me how easy it is to land on Duna with wings, so I changed the challenge. 1.) JebsDead - landed on Duna 2.) Wibou7 - landed on Duna

I already installed the latest one shortly after I made that post, and it didn't have any effect. On the NVidia Experience panel I was able to pull up a page for KSP which showed the recommended settings for my rig. They appeared to have settings from before the 1.1 update but it only told me to turn them higher than I had them, in some cases higher than the new UI will allow. It recommended 64 shadow cascades, for instance, while it only goes up to 8, and apparently due to some other change, even at zero shadow cascades the shadows look excellent and sharp. I checked my CPU core temperatures over several minutes and sitting idle in the VAB they were all 4 running yellow at ~80-90ºC. When I built a very basic craft using no more than a dozen or so parts, the fan spun up even higher and the temperatures rose past 90, getting close to 100. Maybe there is some issue with the heat sink, but it can't be that bad because no other program does this to my computer, including KSP before 1.1. The only thing that even came close is the Minecraft forums which are phenomenally poorly constructed and cause my CPU to spin up to high numbers to load nothing but junk graphics that are not only excessive for a forum but also should load far more easily. They don't know how to write code. But there are no games I play that stress the processor. Minecraft, WoW, Path of Exile, EVE Online, Starcraft 2, Diablo 3, Wildstar, Starmade--all these games run fine at high settings. You might say that these aren't the most graphically intensive games out there, but I would say neither is KSP. I have a copy of Metal Gear Solid 5 I haven't installed yet, maybe I should try running that and see what this rig can handle? edit: Actually I don't have a copy of MGS V because NVidia doesn't want to deliver on their promises. Yet another reason I don't like to deal with the triple-A developers.

I feel like it's not enough to be able to build a base on another planet and bring supplies and craft to it. I want it to feel like an accomplishment when I set up the base. I want it to really provide me with some leverage out there, and not just be some cumbersome fueling station that's more trouble to use than it's worth. Realism be damned, I want to be able to use on-site supplies to build a vehicle assembly building wherever I want to put one, and construct ships right there that I can launch from a pad on that planet. I'd like to see a mod that lets you use some resource which you can either collect on site or ship from elsewhere (including KSC) to construct a VAB and launchpad anywhere you want to have one, and then use more of the resource to construct ships for launching. It should probably be a new resource of which there will be tanks for it that you can launch from KSC already full of the stuff, or you can launch empty tanks and fill them on-site with the in-game ore extractors, by using the ISRU to convert ore to the new resource. Call it Skeel, if you can't think of anything else. Skeel, an ultra-dense metal with a moderately high strength to weight ratio and unbelievably high melting point which kerbals build all of their spacecraft structural elements out of. I'm not expecting to win a prize for the name. Are there any mod projects doing something like this already? And if not, any of you modders feel inspired to start one? =]