Jump to content

king of nowhere

Members
  • Posts

    1,955
  • Joined

  • Last visited

Posts posted by king of nowhere

  1. before launching a vehicle that is ever slighly different from everything you have launched before, save the game. then use the cheat console to test your vehicle on all conditions that you may encounter during the mission, perform all the tasks you may want to perform. no matter how routine, perform them anyway. then launch the mission, and run it all the way to the end. don't worry about anything else, you're going to reload afterwards. only after you succesfully ended the mission, you can go back, reload the previous save, and do the mission for real.

    i don't remember it ever happened that i didn't discover any fatal design flaw the first time i launched something new or tried a new mission. from a docking port installed backwards, to a rescue vehicle missing a parachute, to the thing not having enough reaction wheels, to an ore extractor capsizing upon attempting to pierce the ground because it wasn't properly secured. something will always go wrong at first. always.

  2. 14 minutes ago, AHHans said:

    Do you have a backup of the KSP installation of the saves folder somewhere? Or do you use steam with the cloud-sync?

     

    nope and nope.

    i am old school. i come from the times when each saved game was a backup of its own, so saving the game multiple times was safe. now that i think about it, the far cry games also work in the same way, and i always thought it a terribly stupid way to manage saves.

    but i'm thinking: the folder got overwritten, but the files in it were not. so the individual files must still be somewhere in my hard disk, and it should be possible to recover them using some specialized program. i will ask some of the pc experts i know

  3. i decided to try running sandbox to see how it looks with the parts i'm still missing.

    so i went to create new game, and i didn't bother to change the default name. it asked me if i wanted to overwrite the save, and i said ok, because i wasn't even using the default save, i keep lots of saved games anyway.

    so i check what i wanted to check, then i come back. and i can't find any of my dozens of saved games anymore. nothing. i open the game data directory, look for the saved games, nothing there. i start a new career with a different name to check if i have the right directory, and indeed i see that a new folder pops up. and it's the only folder there.

    does this truly mean that i accidentally deleted all my career? all the dozens of saved games that i specifically made as multiple backups, all the ships i ever crafter and saved, hundreds of hours of gameplay, everythingis beyond recovery? please tell me there is still hope

  4. if you are looking for a challenge about reaching other planets, how about using gravity assists? when they sent a mercury orbiter they used 6 gravity assists between the earth and venus to slow down the ship, you could try something like that. even just an eve flyby with atmospheric braking should save a fair amount of deltaV

  5. I am working on a multipurpose rover, it works magnificently at most of its tasks, yet it has a couple of weaknesses i could not manage to fix; i have to ask for advice

    the rover has many functions:

    1) fuel transfer between ISRU facilities and landed ships.

    2) collects science

    3) can land and take off under its own power on all planets without thick atmosphere (up to duna)

    4) is fun to drive from a first person perspective (hence the cupola module oriented for better view of the ground)

    5) survives capsizing at high speed

    6) can operate in a wide range of environments

    the rover has around 15 tons of dry weight, with capacity for 32 tons of fuel

    https://imgur.com/a/OuInRNY

    it does those things quite well, but it has a few weaknesses

    a) the rover has a hard time moving uphill. the slope limit depends on gravity and amount of fuel, but it's pretty low. fully loaded, on mun it can't take much more than 10 degrees. even on minmus tiny gravity it can't climb steeper than 30 degrees. this makes it almost useless for moho and duna.

    unfortunately, the only ways to fix this that i see would be more wheels and/or bigger wheels. i already put as many wheels as i could fit on that design. as for bigger wheels, I'm already using the TR-2L model. the only model bigger than that is the XL3, but those things are super heavy. even if i swapped out 10 old wheels for 4 new ones, i'd still add 4 tons of weight. way too much. i guess i could use the hinges to angle my rockets backwards. i don't want to have to rely on consumable fuel just to move around a planet, though. and while i don't need it fully loaded, i'd like to be able to move around a planet with enough fuel to take off. which, in duna's case, means between one half and one third of full load.

    b) the thrusters don't give uniform push.

    this is harder to fix because the center of mass of the vehicle changes with fuel load, as evidenced by the image. however, the asymmetry of the push is small, and i was hoping the SAS system could compensate it. indeed, it can at low thrust. my previous version of this rover was able to achieve orbital manuevering, orbital rendez-vous and docking and a minmus landing, despite being more unbalanced than this one. however, taking off from mun is complex, the rover risk spiraling out of control. i managed after a couple of tries. i could not manage to take off from moho without losing control (i could land just fine from orbit). by the way, that was done at full load, when the center of mass and center of thrust align almost perfectly.

    the problem is made worse by the hinges vibrating and further messing thrust direction: the rover is much more stable with the thrusters pointing upwards, but it needs to rotate the hinges downwards to take off (so the hinges are the most unstable right when i need more thrust). I need to explain why i made it like that. Originally, there was only one thruster, and its purpose was to right the rover in case it capsized. only later i realized that since i already has a rocket and a fuel tank, it would be easier and cheaper to strap on some additional engines and make the rover fly than it would be to design a whole cruise module and a landing module and adding them to launch mass. but i still need the rocket to work as anti-capsizing mechanism, so i still need them to be able to point upwards. also, the hinges allow the rockets to be safely retracted to be protected by the landing struts in case of capsizing.

    the original plan called for intentionally capsizing the rover in order to take off (i know, it looks silly :blush:). but the rover will just lean on the side, and it's quite hard to take off like that. so, i had to include the possibility to rotate the thrusters.

    i could try to control the vibrations and asymmetrical thrust with a big reaction wheel. i could also use the stronger, heavier hinges. but i'm quite reluctant to add even more weight.

    the push asymmetry could also be controlled by manually setting a thrust limit for some of the rockets, but it needs to be fine-tuned, and there is no way to do the tuning without firing the rockets and seeing if the push is uniform. if it's not, it's generally too late to fix things. and of course it must be continuously changed to accomodate shifts in the center of mass caused by fuel consumption. so, it's virtually impossible to make it work in practice

    c) the rover still breaks for taking a bump at high speed

    the safety struts work very well under normal conditions. i run many tests accelerating to 30 m/s (which is basically top speed with those wheels, and anyway it's top speed on most planets because low gravity will make the rover unstable; in fact, on mun it's hard to drive above 20 m/s for long without capsizing), and i never had any damage. directly slamming on the ground is another matter, though. in particular, taking a bump in the ground at high speed can break the front wheels, if the bump was steep enough. i tried putting the front low struts to try avoiding this problem, but turns out they also can't support a direct impact with the terrain at high speed. i may just have to drive more carefully around rough terrain, if there is nothing to be done about it.

     

    any suggestion on fixing those issues without creating more problems somewhere else? is my dream of a rover fulfilling the 6 goals even feasible, or do i have to accept mediocrity somewhere?

     

  6. i also have a similar problem. from what i can surmise, when coupling two ships the game makes some small correction to their position to fully align them. on the ground, this can cause some parts to go slightly below ground, or the wheels to become suddenly very compressed, and so the whole thing springs back. will try the advice of tampering spring values, otherwise the only solution i found is save scumming before coupling stuff on land.

  7. 4 hours ago, KrisKelvin said:

    now that's cool! I did wonder how I was supposed to know what height the surface is! duhh!

    yes, that's a good trick. before i discovered it, i tried to use my shadow to figure out my altitude. and of course i could only land with daylight.
     

    Quote

     

    I've been trying to get it down to 15 m/s or less. Sometimes I think I've just managed it and the damn thing bounces and tips over!

     

     

    are you throttling your engine carefully with shift and ctrl? go easy on your engine. brusque manuevers result in disaster.

    once you are close to land, you want to keep your thrust to exactly cancel gravity, so you float down gently. for most landers on mun, that's between 20 and 15% of max thrust. look your speed; if your speed is increasing, you decrease your thrust, and viceversa. you must be very delicate in your commands to avoid overcompensating.

    10 m/s is slow enough that your ship won't break up and explode, but still, the slower the better. especially on low gravity worlds; you can get away with 10 m/s on mun because there is decent gravity, but on minmus 10 m/s will make you bounce and capsize for sure, and even 5 m/s is risky. on the plus side, gravity on minmus is so low, it's easy to keep a lower speed. also, the more stable your ship is, the more you can be careless before capsizing. my aforementioned first lander could have never landed at 10 m/s without capsizing, not even at 5 m/s.

     

    Quote

    Ah.. I just have the two pilots who are already in the ship at launch. Didn't realise you could pick your own pilots. But surely it's me who's flying the ship, not them anyway? Aren't they just for show?

    yes and no. yes, you are flying the ship, you are giving commands. but the pilots can enable some additional functions. the first is SAS, stability assistance system. it keeps your ship standing still. i guarantee, it's almost impossible to do a good job of it without autopilot. once your pilot gets some experience (or you get a module past the basic ones) you can also point automatically in certain direction. in this specific case, you can keep the ship pointed retrograde, so it will always fire against its current movement. it makes landing much easier, because while keeping the ship poiinted correctly can be done with a bit of skill, pointing the ship correctly while also checking and adjusting the speed, and checking the altitude at the same time, while keeping an eye also on your fuel level... doing that all together would require an ace. good thing you can autopilot your attitude and just focus on the engine.

  8. never considered using a joystick. i suppose it might be easier, though by now i'm used to the keyboard.

    however, landing is not so tricky if you take a few measures:

    - have an experienced pilot or a module that can point directions on board. automatically lock on retrograde, then you just need to burn retrograe and you will cancel all your speed. no need to ever press the SDWA keys, ever!

    - have a ship with enough manueverability that you can quickly and reliably point in any direction. adding reaction wheels help.

    - have a ship with a large base, so it will be more stable against capsizing on the ground

    - do things slowly, have enogh extra fuel to not be rushed.

     

    still, landing is not trivial. it's just routine enough that i can't bring myself to do it as slowly and carefully as i should; plus, going down like a stone and only braking at the last moment saves fuel, giving me a dangerous incentive to do things the dangerous way. finally, capsizing the ship is always a danger if you don't land on flat terrain. if you happen to fall down on one of the most cratery parts of mun, it can be hard to find a spot - and fuel-consuming.

    and especially at the beginning of a career, a major problem is that you often don't have the pieces to make a proper lander with a wide base. my first lander was a very tall and narrow thing that could only land, with difficulty, on the most flat of the terrains. my second lander is quite stable and i'm still using the same model today, but it required unlocking new pieces.

  9. as a long term plan i want to make a huge spaceship that can complete any mission without any external assistance. it can move anywhere in the kerbol system (at least 5000 m/s of deltaV with all tanks full, without counting the landers; though i may have to reduce that requirement a bit), it can drop an isru complex to refuel itself on any planet's SoI, and it can also drop a rover on the surface of any planet and get it back, all SSTO (not sure it is possible for eve, i may have to ditch that part of the plan; but definitely can SSTO a rover on kerbin).

    for me it's a very long term project, something to do after i've done everything else. but you may want to try that too.

  10. actually, you can go from mun to minmus and viceversa with less than 100 m/s if you get the hohmann transfer orbit right. it's just a matter or finding the planets aligned properly.

    even more difficult if you want to use a gravity assist. you can get between 200 and 300 m/s of deltaV for free if you get a gravity assist from Mun on your way out, but it will further alter your trajectory.

    what i prefer to do is to launch from kerbin orbit when the alignment is about right (as seen with the intersection markers), with a gravity assist from mun, and then run a course correction manuever as soon as i'm out of kerbin's SoI because the conics look much clearer if you don't have to change system.

    then once you get there you have to make a retrograde burn, just like a mun insertion. for duna you can use aerobraking to save some more fuel

  11. i am about to launch a satellite to moho, and i suppose it will need some cooling, but i have no idea how much. I'd rely on my tried-and-true strategy of using debug console to transfer the whole craft in moho orbit, see if it works properly, before reloading the game and launching it for real. well, i did just that, and i accelerated time, but i got no heating problems. without any radiators. which i have a hard time believing.

    so, is it true that i can send a satellite around moho without any need for radiators? or is there some reason why the test did not work properly? how can i estimate how many radiators i may need without testing by random?

  12. i don't like to go too big, i try to stay reasonably efficient (600 tons of payload? what for? i never needed to launch more than 30). on a big fuel tank, i attach an engine to the fuel tank with a decoupler. i reach orbit, attach to the station, decouple the engine. of course i need a first stage too. and if the fuel in the tank is too much, i remove some of it. i leave just enough to reach orbit. i refuel from my refineries on the moons. i launched a 60-ton tank (well, a ship made of multiple tanks) with perhaps 30k :funds:of funds

  13. 1 hour ago, Streetwind said:

    As for your issues with filtering out old craft - don't use "relay". The game will automatically use that for probes that have a relay-type antenna on them. Instead, if you don't want to use "debris" for whatever reason, use something that pretty much never comes up. Like "base" or "rover" or "plane" or "lander" - there will typically be at least one of those you really don't ever need for anything else. Even surface bases can be tagged "station" in a pinch.

    but i did use something that pretty much never comes up. and that something is "relay" :D

    debris, i generally explode whenever i have the opportunity. i used lander for the sky crane, and anyway the game will call some stuff a lander authomatically too. i don't use planes, but i may.

    relays? even if i were to actually make a dedicated relay satellite, it would just have to park in its orbit doing nothing. hence it would not create conflict.

    they should add the "inactive" category. or better yet, give you the option of creating your own categories. maybe it will be possible when they release ksp 2

  14. the good old times when i would launch one mission at a time, and do nothing else until it was done. i wouldn't even use the tracking station back then. but as missions gre longer, i started to launch more of them at the same time. and things escalated pretty fast...

    now i have 14 active missions that are being pursued by some dedicated craft. 12 unmanned probes, 5 manned ships, 3 manned space stations, 2 moon bases with a third along the way, and a handful or rovers and service ships that i'm not counting because they are currently docked to something else. a total of 11 ships are scheduled to have a manuever at some point in the future, and more are simply waiting for something more long term.

    and it's becoming increasingly difficult to deal with all this.

    to deal with the crowding, i started by renaming every probe that finished its missions as a relay. this way, i can deactivate the relays and only see active missions. still plenty of those, though. and there's the added complication that sometimes, upon undocking, the game authomatically classifies some ship as a relay. I almost missed one, which i only found back because it was on a short-term mission that i remembered about.

    I started an excel spreadsheet where i detail stuff. for every mission that i'm not going to make in one session, i write down the important details. deactivated probes are written down, along with fuel left and science equipment: i packed enough extra fuel that i may be able to recycle them some day (i already discovered one that i could squeeze some more science out of).

    I make sure to have every active mission with a planned manuever. even if the manuever is 0 meters, simply a planned node to remind me that i should go check on that mission later.

    and then i have to sort the manuevers, see which craft is the next one where i have to perform something. this would be rather trivial if i could sort my missions in order of planned next manuever, but if such an option exhist, i did not find it. so i have to look the big list on the tracking station to see which one is scheduled first. Hoping i don't have two critical manuevers planned at the same time. when i do, hoping that one of the two can be delayed or anticipated a bit.

    and in all this, a single distraction, a single forgetting to set up a node before moving to another ship, or not noticing that the game has automatically renamed one of my ships as a relay, could cause me to miss a mission. lost. forgotten. buried in red tape.

    I don't expect there to be some holy grail to fix this mess, but i'd like to hear your strategies to cope with too many missions to track

  15. 22 hours ago, coop2168 said:

    If you want, do Sandbox so you can get your feet wet and get used to the mechanics of KSP, do career once you are used to the game. 

    i actually suggest the opposite, starting with career before going sandbox.

    the reason is that too much choice is daunting when you don't know already what you want. to a master painter, a blank slate is ready to be filled with whatever masterwork he can conceive. to a novice, a blank slate is just that: blank, empty, meaningless. he'd much rather have some shapes already drawn that he has to color or add detail, because that gives him a clear objective.

    i started with career because i wanted to start with few, simple pieces and gradually expand as i mastered the game.

  16. 1 hour ago, Spricigo said:

     

    I know I will sound condencendenting but:

    I understand that is how you perceive the whole situation, but it is not, by any means, a problem with the navball. It is just a matter of technique that you didn't developed yet

     

    not only you sound condescending,  you sound like one of those old geezers that are all like "we didn't have all those fancy stuff in our days, and we managed just fine". (no offence intended)

    sure. of course you can dock a ship with just the navball, and perhaps some eyeballing. just like you can light a fire with just a flint. and you can even do it quickly and reliably enough, once you're practiced. it doesn't mean that there aren't better ways.

    I can do a docking just fine with the tools i have. i get most of them on the first try, if at least one of the crafts involved has decent manueverability (btw, you also sound like you've seriously underestimated me). that does not mean that i can't look for better ways to do it. i would never want a mod to do it, but wishing for a first person camera to better gauge the alignment and distance in the final few meters is not cheating.

  17. 6 hours ago, Spricigo said:

    No, it wouldn't. The navball is always oriented the same way your controls are, no matter what odd viewpoint you need to look at the docking port alignment/separation. It's just that you are a new player  not used to the navball yet, you still look at the movement of your ship when you should be looking to the navball, you see the target "above" and you when you try to move "up" it goes "sideways" because the camera lies to you.

     

    i do use the navsphere and i routinely perform complex docking. i don't use the rcs system, so when i have my prograde indicator straight on target, i know i am on the right track. i may need to stop a few meters earlier to realign, but that's all. the navsphere is good for that kind of movements. accelerating when pointing prograde moves me towards the target. accelerating retrograde slows me down. and i can fix directions with the attitude.

    but when i try to use the rcs and the thrusters are moving me sideways, the navsphere does not help. i accelerate at a 90 degrees angle, now my attitude remains the same and my speed moves somewhat to the side. how much? how fast am i moving to the side, and how fast am i moving forward? i don't know! oh, i could still get those information out of the navsphere with some trigonometry, but as far as deciding which way to go, it's worthless.

    say i am docking at a large space station, so i don't want to turn the whole station, and i am off center to the docking port. or perhaps i got a mission to grab a floating piece of junk with the claw, and i need to align it perfectly. i would want to move a couple meters to the right, then realign to the port, and then complete the docking. that's the kind of manuever i would hope the rcs could help me with. but it fails, because i can't judge the whole "push a couple meters to the right, then push back to stop". not in a useful time to actually act on it. by the time i figured out how to stop, i am more off-center than before. with an astronaut's jetpack instead i can do it easily, because i can see where they are going. when manuevering a ship, the ship itself blocks my view.

    getting to within 5 meters of the target is easy, once you get the hang of the navsphere. but those last 5 meters would be judged more easily by eye. and you don't need to be buzz aldrin to be able to point to something that is literally at arm's lenght by eyeballing it

  18. driving rovers on a low gravity world always has a risk of capsizing. so all my rovers are built with a mechanism to put them straight again.

    my first rover was a small one for exploration and it had two mechanical hinges that could push it off the ground (it still ended up running around 800 kilometers on mun to chase objectives, though).

    my second rover was made to transfer fuel between a fuel production facility and various shuttles that could land nearby. it needed to carry lots of fuel, so it was much bigger, so the small hinges would not work anymore. the largest hinges are very heavy, so i figured, since this is a fuel tank anyway, it would be easier to just strap a terrier engine on one side, to provide enough angolar momentum to turn it over. Nice! however, being big, landing it was a serious problem. it got stuck in the landing gear of the sky crane, lots of problems involved.

    i was studying an updated version for minmus. for start i transferred all the science equipment on top of it, so i don't have to use two rovers. science equipment is light anyway. then i was studying ways to land. i simulated a landing with the sky crane, it was doable but hard. the normal coupling point did not work, and the alternative coupling point caused it to wobble too much.

    then i realized: i already have a rocket on one side, powerful enough to lift the thing in mun's gravity. I may as well strap a rocket on the other side to balance the push, and just land the rover on its own power.

    now, since i have a rover that is a big fuel tank with small rockets, i can also use it for the cruise; i only need a rocket to boost it into orbit.

     

    And so it was, that i accidentally designed a rocket car :cool:

    earlier today, if somebody had suggested i built a spaceship capable of interplanetary travel that can then land and move around with wheels, and even take off again if the world is low gravity, i would have laughed and told them it was a ridiculous idea.

    i put in a command module because, while the rover can work remotely, i wanted it to be able to carry around a scientist, since it also works as science rover. i decided to pick a cupola module to give the astronaut a good view outside. then i angled it down a bit to have a better view of the ground. now the rover can also be piloted nicely with in-cockpit view. it feels good to be able to drive a rover with a first-person perspective

  19. even with the error for gravity, though, the difference between the calculated 5600 m/s and the VAB nominal 666 is way too big. using the correct value the answer should be around 3000 m/s. which, by rule of thumb, is what i would expect from such a setup anyway. two thirds of this rocket is fuel, it ought to be much more than 5600 m/s.

    so, there is still some problem here. probably with the rocket, maybe there is no fuel transfer somewhere or something like that.

     

    by the way, here i see another mention of delta-V maps. So far I just started to dabble in outer planets exploration, and i don't really know how much fuel i need to get to the other planets (my solution is just to pack a lot of deltaV in the transfer stage). it would be useful. where do i find such maps?

  20. On 6/13/2020 at 8:58 AM, Fraston said:

    Have you tried docking mode with regular RCS? It works like a charm, since you aren’t doing much orienting, just moving on a 3D plane.

    i just tried the docking mode, but can't seem to get the trick.

    the problem is that i still need to do some orienting to fix the proper orientation, especially when grabbing something with the claw. and docking mode does not, that i can find out, let me fix my orientation. and it seems to create problems there if i try to switch back to normal mode to change orientation and back again.

    also, it's exceedingly difficult to estimate the orientation of your craft and the subsequent direction of your push. the navsphere only helps so much. a first person perspective would actually be more helpful, but it would force me to send pilots when a simple automated probe can do the job for a fraction of the cost.

    in the end i did a lot of swinging around fruitlessly, until i decided to get back to the old way i know, and i docked more easily

  21. sometimes an automated probe stops working properly. i assumed it was for some sort of signal loss, as it generally happens in mun orbit in places where signal loss would make sense, but this time i cannot explain it in any way

    https://imgur.com/a/wfI1Z7G

    i am in low kerbal orbit, i have level 3 tracking station, i have plenty fo electricity, a HECS probodobodyne. not to mention a couple dozens satellite around the kerbin system, each one carrying a good antenna. i checked the antenna and commnet entries on the wiki, and according to everything written there i should have perfect communication. i do have perfect communication according to all indicators in game.

    and yet i cannot control. i cannot control my attitude on the navsphere, and i cannot throttle my engine. both things happen when i have partial communication with the probe, but this is clearly not the case here. (and, as you can see, it's happening at the very worst time).

    so, why am i losing control, and what can be done to avoid it?

×
×
  • Create New...