PB666

I got my 500t ION Driven Cargo-d lander to Moho (and 12 hours of burn) with a residual dV of 33,000 only to find I did not put MechJeb on my lander, and the struts are not designed right look right, and left 6 of the crew back on Kerbin.

I think the possibility issues with jet and LV-N are limited in reality because of the safety issues. NASA could save a ton by placing their launch site on top of kilamanjaero, but then how feasible is it to place stuff on an equitorial volcano. The LV-N is a risk to everything on the eastern seaboard of the U.S. and can you imagine how risky is would be to launch space craft with jets as the launch stage. Not everything that on paper looks efficient proves to be efficient in its implementation. Consider highspeed rail versus airlines, in theory highspeed rail should far outstrip an airline in terms of profit, in reality HSR with an abundance of passangers tend to break-even or loose money relative to airlines.

This logic has problems, to optimally exploit the mun requires 2 cap, 2 mat st. and 2 goo per biome, if one does the mat to optimize sci. per landing, and opts for 2 or three hops per landing then the weight of the LV-N is less critical. 48-7S is great for its weight but its basically a flag-planting engine (best places to land and need to plant flags for their science lander to follow), but when the payload is also heavy, you really need ISP. The problem LV-N in science landers are the non-payload parts needed to stabilize the beast on the ground. I got around this problem by making a pancake elliptical fuel tank the superstructure of the craft and placing everything bilateral of the engine, scale modded MGS for gear. If you go for an out-of-the box lander, then you many need 15 to 30 missions to get all the sci from the Mun. If you think outside of the box you can do the mun in 5 without needing a rover or carry around science stuff on your back. Then you simply keep one kerbal on a site to plant flags for $. Plant and pull, plant and pull . . . . . .

Not specifically that glitch, but similar glitch with other part mods. You've got two parts with the same name PART { // --- general parameters --- name = mk2whatever <------Make sure that no other part is using this same name module = Part author = . .. . . simply change the name of one to mk2whateverAlt make sure the alt version is the non-squad variant.

Squad has been moving parts (renaming part files and moving the folders where the parts are found) around in the folders since 0.25. If your mod depends on a texture from a part that has been moved, then that texture will show up funcky. Many of the mods of fuel tanks rely on stock textures and repaints. They will not work properly unless they are installed precisely and the stock parts location is conserved. I have had to trash many mods from 0.18 and 0.22 for this reason, and my 'reload' log is full of red-lines, theres alot of stuff I still need to throw out. For the most part, however, these mods are now obsolete, I keep them only because of the cfg coding examples.

Couple of issues, on my 0.90 install the program reset my screen resolution to a scale that was distorted, I had to tinker with it to lower the resolution to a scale that optimized performance. Your core processor is a bit slow for the physics stuff some folks are having a problem with. I heard that 8.1 and 64 bit are problems. Make sure you are running the 32 bit version of the game. I have also read but cannot confirm that KSP likes to have its own dedicated processor in multi-core systems. This argues for machines that have higher processor speeds. Make sure your modules are all up-to-date, some modules may have problems in 0.90 and need to be shut off.

lol, but the 909 should be called the lap-dog due to its rather shackled performance. I've named my engine mods also I have the: Mustang, Schnauzer, Miniature schnauzer, Minotaur, Maxotaur, Nutwarmer Atomic Rocket (due to its tendency to leak massive amounts or radioactivity). The Mino and Maxo are rescaled and tweaked 48-7S. Better back my gamedata and parts folder now. - - - Updated - - - Yep, but occasionally you need that specialized part otherwise your millennium falcon will end up looking like a piece of space pizza or as others say 'a dogs dinner'. Lately I have taken to modding the MK1 tank into various shapes because it has simple shape and texture, modding the other stock tanks often ends up not fitting right. MK1 style looks more realistic than the later version of the tanks. In this case simple is better. Before I started reshaping the MK1 tank I had alot of dog dinner stuff. OTOH, the MK2 spaceplane parts are an excellent example of a specialty design that open up whole new avenues of creation, the mods I've made of this series are limited to resource replacement and attachment node rotations. The problem with SPP is that if you go about re-dimensioning the parts they will look funny and not fit right to anything, but......if they perform their function so well to begin this, then that really isn't an issue.

This is the same ship I have launched several times. However this time I closed the game and restarted and the strut connectors did not break up on part separation. The payload is hanging about 0.3 meter over the launch, and the decoupled drifted 1000M away and is strut connected to tank (and I have no idea how it became connected).

It may not be missing, you may have accidentally moved so far to the edge of your screen you can no longer see it. There is a way to reset the game environment that involves erasing a game file and forcing the game to rebuild it.

Whoever can figure this out gets eighty bucks and a hooker. Thank you

2.5M wide tank with landing struts barely fit, once out they may not fit back in. Good idea however, using the cargo bay to deliver rovers.

Sorry I had to fix this for you "Kaetheist". He did use the word seriously and who does believe works of fiction? Thread is cooked, stick a fork in it.

Excellent example! But wait, if you are a purist how is this not an exploit, If I make a really tall rocket I have to build XL MGS and then strut to other XL MGS to stabilize this. Think about this, I can now use these interstage struts, make then 10 times stronger than the EAS-4 and by pass alot of parts, lag issue, etc. (BTW I stated in this forum that the strut connectors need to be refined, its time) In the end the outcome may be different. In the stock part case I may abandon tall rockets for short fat things with side tanks and asparagas layouts that lack aerodynamics. In the mod case I make them taller and sleeker like they should be. Basically you are arguing that 'feel good about bypassing game limitations in one direction is OK but not the next'. Here is precisely what I do. Factor 5 or 6 tank, lots of 4nozzle rockets on oversized radial attachment points (Because I know how drag is factored and I really dont care about drag and but do care about part count lag -avoid the kracken at all costs!). Plop my payload on top and strut the radials (Nacells and other lateral projections) to the rim of the tank, SSTO. Its no different from using unlimited fuel. With two stock mods I can essentially achieve the same result as unlimited fuel. Even a mega tank launch has limitations, I have put a 106 crew capacity circular-spinning space station in orbit the limitation is the width limit of the launch pad. I put in on a rocket for the sole pleasure of watching it launch (or not). I could have just hacked gravity until it was 80Mm and done a 100 dV burn to orbit. Aerodynamic rocket parts (aerodynamic cones) offer no game advantage, they are a disadvantage, the conversion parts (unfueled) stack 3.75 to 2.0, 2.0 to 1.0 and 1.0 to 0.5 offer no game advantage, they increase weight, increase cumulative segmental flexibility and increase drag. So if I create a modded tank that corrects in game deficiencies should it not offer up advantages of having thought about how the game is unrealistic? Yes it should. And a few months later, here comes the Squad version of the same tank. Hmmm, immoral or simply ahead of the game. Its not even grey zone, its all white. I have no problem increasing ISP of a part that is 50 year old technology that was never implemented, If I am going to use that part in a specific application (e.g. a NERVA like rocket that is going to be centerpiece of a low-g lander) and efficiency and reuseability is going to be a centerpiece (paraphrasing the wiki on NERVA, it was designed for hours of continuous use). I read the real-world stats on all the engines I mod, and I try to keep those stats within real-world limits. If someone takes my mod (although I have only presented one as an example of node shifting) and buffs it up with ISP, I am not sure why I would care, I have buffed up others parts, played with them, I have a whole directory of mods I don't now use but I reference the coding for future constructs. Even the bad mods I really appreciate because sometimes they give insights. Art is that which expands on the boundary of art. But the critical point is that knowing about the game deficits and thinking about game deficits is - by definition - part of the thought process that will create exploits, even if I don't create them myself, as soon as the new stock parts appear I am ripe to include them more so than if I had never tinkered with the beast.

Saw it and chuckled, how the internet has changed since 87.

If your a purist, yes. In the abstract, every morsel of information learned bout the kerbalverse lends information on the when and how of creation and ultimately the utility. A universe that is 6000 becomes 13.3 billion and for some, multiples thereof. Each Kerbal scientist adds a bit that takes power away from 42 and gives it to the kerbals, eventually they can create their own Kerbalverse (e.g. Kiggs). In a medieval religious perspective, it was not just the act of ..., but the carnal knowledge that was the 'sin'. In the same way every morsel of information one learns about the game engine and algorithm alters the way we design and play. Its a Genie that cannot be shoved backwards. You cannot argue that a modder is a lessor player simply because they are more informed about game play mechanics. Better or worse at docking, transfers? Before using MechJeb, after using MechJeb. Another example, I have 4 strength variants of strut connector, using these I learned emperically about where designs weaken, thus now I use far fewer struts by making designs that are inherently stronger. I could remove the strut con variants, but not the knowledge I gained in designing and using. One final example that will add sense. If you are playing career hard and have to go to the persistent save to uncrash a game, simply knowing of the file and its content is abstractly god-mode, for me the game is over. To be clear, if Jeb falls through a hole in KSC terrain mesh and dies, and I go to the persistent to rescue him, whats the point of career hard mode. Hard mode and any form of persistent save manipulation is a contradiction in desired imperatives. Either the engine has to work perfectly and be opaque, or career mode is a compromise, and I start thinking about its design (e.g. why manned flight before unmanned)............. its done. (am I an career mode ultrapurist?) But in the creative mode no such 'ethical' limitations apply, simply stated I can have as much knowledge of the rendering as I desire and it only adds to and improves the game. In fact I personally would like to be to the point were I can Blender a part, create a collision mesh and port it into the game as easily as I mod stock parts.

Today, the issue of game morality has spiked this forum with all manner of concerns. Noting that the Promo of the game is two Kerbals grinning at their overturned lander "Mun or Bust". And taking in account: rumors of earlier versions in which screen shots of a launch blowing up into a great number of radiating fuel tanks was grounds for congradulations and laughter in the community (And xkcd cartoons); The word 'Kerbacidal' or similar variants has made the national news media. (From the news bite meaning not caring that a player has wiped untold hundreds of Kerbin-kind in all mannner of launch, land and lost of control disasters). Given the fact this is a single player, heavily modded game, in which modding is so integral to the game that current mods may end up in the next release, and Curse is a sanctioned modding center. . . I can dispell the morality issue simply by saying KSP is a form of interactive art, archetecture or engineering, as long as the creations are not traded for profit, its amoral. You are the creater and consumer of the art you create, and art is as such until you market it as something else. But if we must discuss ethics of play . . . . . 'Alt-F12 is cheating! You're a cheatah'. Uhmmm, alt-F12 is a place for loading new and edited parts. Its a place to debug and look for problems. Its a place to tweak a resource or the physics so that one does not have to create a 1500T launch for a payload that will be unstable to time warp once it reaches orbit. Alt-F12 can be used for 'cheating' but I suspect most use it for modding and testing parts. If the game is a creation and alt-F12 (a part of the creation process) is 'cheating' one can only conclude that creation is a form of cheating? (I will deal with the broad philosophical context of this thought at the end). If the game itself is a consequence of 'cheating' and cheating bothers the player then probably shouldn't play? Might I recommend bingo. 'Unlimited fuel is cheating', against who?, but using unlimited fuel to get to Moho removes the challenge of the game. Think about it this way, if career mode is the only true game mode, then all other modes are cheating, but prior to career mode we were all cheatahs. Part Clipping is a 'Sin'. Use that at your next confession. So My understanding is that the reason that mass models drag is the complexity of a surface based drag system. In a real world 'milk' . . . . . (I don't see any cows on Kerbin).... In a real world deep space kracken did not rip your ship apart because of the game-clock time differences between a part A and close part B had to, during translation, factor in delta-d/delta-T as T shrinksto the tick in the game clock. Space was very big and numbers could not deal with it all at high speeds, Certainly that makes a difference at the horizon of a black hole, but. . . . . OK so my understanding of clipping is as follows, in old game, you crammed parts and overlaps parts. This was IIRC a game issue, not a physics issue. When you move fast and they impact each other at a speed higher than their impact tolerance and they explode. So point centered mounted them on meshes and made sure they stayed clear of each other using struts, and the folks in the 'help' forums have a much easier life. This is my understanding of why 2 versions back we avoided clipping. In a real world parts would be hidden or molded into each other to provide the sleakest structure as to minimize drag. If we want to argue real-life, the perfect structure for surpassing a ground speed of 240 m/s (~Mach 1) is a fusiform shape, along the ray of attack (any number of parallel lines representing the angle of attack) there should be a smooth rise, flattening and fall of crossectional area. Sudden increases in crossectional area (such as a goo container) can cause a rapid build up of density. In speed photography these density modes are often trailed by a momentary cloud(s) of 'steam' as the rapid change of pressure causes momentary condensation of water vapor. The critical point is that density rises astronomically along surfaces as speed approaches Mach 1 at increasing angles to the 'ray' representing the angle of attack, and in real life this has had disasterous consequences. Below 23,000 meters how many times have we(pl) surpassed 240 m/s, most of prolly cross around 10,000 meters and note that the jellyfish shaped asparagas did not explode. That is not real-life, that is _game_ physics. IN real world sticking out non-aerodynamic stuff is kracken fodder for V1 rocket and 50s era Jet craft alike. Stock parts purism. I could make ton of reasons for modding. Keep it simple. I made two adapters (fueled) in 0.25, the next version 0.90 has one part that does exactly what my two together do, based on the purism argument the stock parts would be a 'cheat' or 'exploit' of my two part adapter. If your are a real purist you prolly should not use any parts except the MK1 fuel tank an the basic engine and a retro-MK1 capsule. Modding IMHO, is the only good reason to keep playing after completing the career mode (or previously sci mode). X is that which expands on the boundary of X. This equation differentiate abstract thinkers from everyone else. Physics-less parts are a 'Sin'. Really, you are not going to use a fuel-line or RCS thruster because its physicsless? Evolution.....of everything in the Kerbalverse. Koses - all was created by that which exists it was. I have faith 42 did it. Kewton - matter cannot be created or destroyed, energy cannot be created or destroyed. E = 1/2 m v ^ 2 Keinstein - there is an equivilancy between matter and energy through this equivilancy they interconvert E = m c ^ 2. Also gravity and inertia are related because energy warps space time. Keisenberg . . .Klanck, there is a scale to the universe at the lowest limit matter and energy can spontaneously be created and destroyed in small discrete quanta (e.g. matter can be in multiple places and no place all at once). E = hv on planks scale E = v = m h = C = G = 1 See what happens if we start to question the system . . . . .I can summarize everything....42.U chaos. MK1 capsule disappeared..a longer summary? Kewton challenged 42, taking offense Keinstein cheated Kewton by stealing energy at the speed of light (a crafty slight of hand that also curved space, clipped time into space and modded mass and energy). Noting the favoritism Keinstein showed toward 42, Keisenberg tricked 42 into playing dice with the Kerbalverse by getting Kawkings to convince the Kerbalverse that it was a quantum singularity on Klancks magical scale. That shrank K so small that 42 could not see it. Not to be cheated, 42 imbued Klanck's scale with quantum uncertainty (Also known as ALt-F12). Utterly dazed by the fact 42 would play dice with the big (oops, small K, damn that scale modder Kawkings) and therefore forever plaguing statistics with false positives and negatives...in less than minus 42 of a second K inflated into a chaotic state, thereafter converting most of its physicsless parts into draggy massive things (many RCS thrusters became "Mainsail engines", however, most just exploded due to the fact they self-modded into parts that lacked of impact tolerance and force resistence). For the next 200,000 years (or 200 or 20,000,000 depending on your temperal reference frame) the false positives and negatives battled the truth in a process called annihilation which left K hypnotized (Also the old MK1 capsule apparently was annihilated). When K woke up it was big and growing but why? And it had a hangover we now call KMB and doesn't remember a thing about what happened. There may be another kerbaliverse in it's future explaining its large size, since Kiggs apparently took liberty with the little K while it was in its' vulnerable state, and spread little swimming bosons in that space-time where-the-stars-don't-shine, and any significant bump in the field may mean a baby Universe is in our future. We must clamp down now, or K may start on the slippery slope toward toward energy and matter 'from the dark side', oh wait, too late. My two edicts in career mode 1. Moho is the only planet of significant meaning, once you reach Moho orbital Pe, with enough fuel you can reach anywhere, easily, quickly you may not be able to stop once you get there, but thats another problem. There is always enough energy around Moho, all you really need is something to propel. 2. Jeb is not kracken food. Once you feed the kracken the game is over. Nobodies perfect.

In the beginning there was a dimensionless singularity within Planck scale, inflation . . . . . .cool, charge neutralization . . .accretion, evolution, Big Bang Theory, KSP, this thread . . . .Finally. .then it exploded

Done it, easy, just make sure you correct trajectory within an hour of landing as to target a meteor crater (or have good wide-set gear). - - - Updated - - - This one, I was about 2 months into the game before I realized that the decoupler had an arrow on it, lol. After the 1st week I only used separators and it took a long time to trust those decouplers again. - - - Updated - - - MechJeb1. Remote Guidance Unit (Physicless) = justification enough to use on all jettisoned utilities (such as a fuel tank with a dock that can be used by craft to refuel) tank, solar panel, dock port and MechJeb = refueling station. 2. Much better in docking guidance relative to stock. Target mode +Target and -Rel. The speed of adjusting target orientation relative to manual all by itself is noteworthy. 3. Suicide burn count down. Enough said, even if you know everything there is to know about space flight, you can still benefit from these utilities Add to these. . . . . Precision transfers because you can precisely match orbital planes (+/-NML feature). So for example you can hit the SOI of a floating snowball called Eeloo close to Apo from an orbital burn out of Moho without having to make a single midtransfer correction.

If you are playing career mode hard - its a game If you are modding parts and creating new craft - its an art If you are testing those parts and creating parts for new utilities or testing physics such as atmosphere curves and delta V, then its a simulatory The answer is not black and white, its red green and blue.

LV-Ns work well on non-atmospheric planets. I have kerbin launches that combine LV-N with SRBs. Halfway through gravity turn they're a competive choice. - - - Updated - - - I would say do a Duna flyby because you failed to consider that stock ION drives lack sufficient thrust to retrograde to a planetary orbit.

So many like questions today about play morality Alt-F12 unlimited fuel Alt-F12 Hack Gravity Can be used for parts testing (payload) in sandbox mode and I see nothing wrong with that. I have used alt F12 to explore Kerbols corona (see what the death radius is) and similarly what are deep space limiits. Use it, test parts, build better, improve.... This is not PvP in WoW.

Agree. Long live the Kobiashi Maru E = m c ^ 2 also known as the mass energy equivilance. in Natural units in which C = 1. E = m. According to Hawkings and based on the standard model of partical physics the Sum(E) occupied on natural unit of distance (plank's point) (The dimensions did not exist). For at least on plank's unit of time. Fermions determine occupancy, if we convert fermions to bosons we can keep an unlimited quantity of milk in a gallon of milk. Getting the milk back, of course, is the problem. In Kerbal a gallon has the density of 8 gallons, and a fuel tank is partially occupied, therefore there is considerable room to merge.

In 0.25 i did not lose a kerbal. In 0.90 I have lost Jeb 3 times a walking to the grass kracken in KSC. Mmmm-mmmmm. So if you are saying play but don't walk then that might be correct, but I think EVAs are part of the game so.....

Having given up on career mode for the time being . . . 1. Way too buggy, even with the work-arounds 2. Done it in previous version 3. The Astronuat complex needs a golden suicide pistol mounted on the wall, pretty much this is what career mode is now. 4. Why exactly do we start career mode with manned missions? . . . . . my attention has shifted to parts and sandbox mode. My first impressions here were ... Hey I modded a part that almost just like that. Still keeping my old mods but the new ones have found their uses. I have done a fair share of new part modding, in particular the Space Plane II parts. - Dupped a SP2 crew and made the primary node a back attach node. - Dupped and Shrank the F2 to F1 fuel to F1 to F1/2 - Dupped smaller conversion and gave it 'special' properties Rant- Finding the SPP directory was not easy - shouldn't it belong in the parts directory? From these parts it was possible to build a sleek 2 @ 2 symmetry space ship with a 4 nacell design really worthy of sci-fi notoriety, Bravo on the new parts. Comments: 1. however, the capacity and fuel density of the SP2 monopropellant tank is about 2/3rds of what it should be. 2. Another bravo goes the the considerable number of fueled conversion parts, I found one deficiency, but more or less every conversion is covered. 3. I was looking at the SP2 'telescoping' docking port, and I wondered could you actually transfer crew through such a crapped space. My solution as to convert a Side-Adapter mod into a back of the part as crew and goods transfer area. 4. Unrelated to new parts but somewhat related to 3, I think a factor 0.625 to 0.88 basic crew transfer shaft would be a nice addition to some ships, something about as long as MK1 but half to 70% as wide. 5. I am looking to build parts for a high ISP VASIMR-like ION drive, there are issues discussed below. The ship was designed to be a true full feature interplanetary ship, capable of performing a multitude of independent missions in one trip, in this it accomplished its task but with exceptions. 1. To conserve fuel ION drive was used for the interior planets, it was decided that since a leg could be made from Moho to Eeloo taking full advantage of the Oberth effect and the proximity to the Kerbol this was a good choice. 2. To utilize ION drive as frequently as possible on the burns including the landing burns (this is possible on many satellites). 3. Refuel as needed. In the trip to Moho it was quickly realized that the mass density of Xenon is really low, one either has to bring a hideously large and wide Xenon tank into orbit or a bunch of F2 tanks (big-orange '64') dimension and chain them together on docking ports. We all know what happens when we do this [Kracken noises heard in the back ground]. The only other option is to use a F3 docking port. The SP parts work so much better than previous parts in terms of stability. Despite having a thrust slightly above the S gravity of Moho, the combination of high ISP engines and ion drive got the ship down safely. Getting up was no problem, but on the Moho-Eeloo transfer, Xenon was a major issue. The solution I have decided for this multimission ship is a higher ISP ION Drive, otherwise one needs to cheat on fuel density. There is plenty enough energy around MOHO to justify the change. This is only a ~5500 D/V and with high ISP (2000 to 5000s now) engines is should be easily accomplishable, except for the feul density issue. Eeloo landing - a breeze, no problem. Eeloo to Jool transfer - surprisingly easy (although took 5 years this ship has 7 crew and space for 42). Jool - Pol - 1st major problem, ship was damaged on landing - flight engineer I brought along was useless. (see below) Jool - Bop - Big problem, this is not the place to land a 100T (empty) ship. The biggest problem is that there is no unambiguous reference for terrain. I find myself saying 'this l o o k s? like a slope' to move over and find what looked like a 20% grade from above is a 60% grade from the side view on descent. The only real positive reference was the asymmetry in light spot created in the 'night' landing cameras. This can be exaggerated by having two of different colors one mounted further back on the ship. However this only works well if the ship is within 100M of landing, not the time when to decide were to land. ON Bop there is the simultaneous issue of getting close enough to survey, changing direction, and staying far enough the terrain to avoid collision. Without telemetry, for large ships one needs to send out probes with rovers to select landing spots, this creates another problem (see below). Terrain Guidance What we need here is terrain guidance of some sort, for example a switch that shows were the lowest and highest points in the terrain. Alternatively another map mode in which the gravity centered point-centered mess can be brought forward or backward against the terrain to see where the mesh first and finally breaks the surface. The irony here is that I have a massive space ship with a crew of 7 and 2 dozen veiw ports but still have no radio telemetry capability, even 100 M off the ground! The Cargo bay objects and lag. I have to say from the start that the new 0.90 parts have greatly simplified ship building in terms of parts and need for converters etc. A comparable ship of complexity 0.90 versus previous 0.25 has lower lag, but with the new capabilities also comes the new sources of lag. I wanted several probes that could be dropped instead of landing. The weight and size was not an issue, but each probe adds to the part count. I was wondering why could not parts be frozen in a SP2 cargo bay when the doors were close and the part count diminished to 1 for the cargo bay only altering the weight of the cargo bay. I thinks a real good idea to mask part intricacy while the cargo-bay door is closed and SP2-cargo bay is structurally intact. Basically 4 small probes (F-half) slowed the game down from just under normal time to about rate = 1/3 rd. The second issue was I fixed my sat with a modified smaller radial decoupler to the back, should the decoupler not be active until the cargo bay door is open. What Iam trying to say is that should not all the activities of parts in cargo-bay (e.g. fuel tanks, parachutes, etc) 'silence' while in a closed cargo bay? The strut connector needs a more civilized big brother. With the advent of the new VAB 'tweak' tools, the strut connector needs to be appended. I Use a MK1 that has been side flattened to form a nacelle arm, it can be attached via strut connectors after tweaking, it would be nice to weld to parts together once they are inside a certain range (say 0.05M). Flexible attachment schemes. The use of adapters such as the radial attachment points (which I typically use for mounting complex engines to fuel tanks) are not really necessary under most uses and add to the part count. There should be part attachment nodes that can be 'stuck' in multiple, user chosen, ways. For example if I want to node- stick an engine to a tank, or stick the same engine in a 6:1 configuration to the same tank, I should have the option. Finding stable places to put nuclear engines and ION drives is not easy, and most endup cranking up parts count, hard to get into space, require unnecessary and unattractive strutting and a kracken fodder once there. These problems are largely solved by getting rid of the need for 'bridge' parts and creating flexible attachment parts. Appending the above: Cargo bay wobble. This goes to the suggestion above that activities of objects in the cargo bay should be silenced. I found my cargo-ed lander was wobbling excessively. If I silenced the reaction wheels elsewhere it stopped, then finally silenced the reaction wheel on the lander, and turned on the main wheel.

Over 100t, using scale modified stock parts. The lander I am currently building is 124T and has a capacity of 41. It will only land on celestials with g less than 0.15, but can to atmospheric science on some planets without landing. It uses a collection of VASMIR, and SSME like thrusters. I'm guessing it will get 105T back into orbit. For Mun like celestials it needs to be refueled on the surface. I have landers that could land and take off from 4 different Mun craters before needing to refuel in orbit. I used a single lander in a cycle of 4 refuelings to collect science from all the biomes of the mun. I'm guessing its mass was around 50T. At its core was a scale modified NERVA-like thruster, The entire ship was build around the engine (literally) and scaled down and rescale parts inorder to optimize storage a minimize mass.