Kertherina

How are you supposed to land on Venus? I made a successful entry just yesterday and even successfully landed in previous versions of KSP / RO, but now once i reach 4000kPa pressure everything just explodes. Since as far as i know there are simply no parts that can withstand more than that (i've looked through most of the tech tree), how can you make things not explode before reaching the ground? Should you simply not use the option where pressure can destroy parts?

I occasionally have the same problem and also KSP crashing when trying to load crafts, although i wasn't connecting them before. Also using RO and RP-0.

Thank you very much, one of the mods i really missed with how impatient i am, especially during reentries

I don't use CKAN so here's a screenshot of the GameData folder: http://i.imgur.com/oC3koi2.jpg (The Mainz_EngineAttachRules.cfg allows for surface attachment of all engines, not sure if relevant but just so you know what that does. Also, i think i added KJR afterwards)

I also have a problem with engines specifically breaking way more often than they should. I basically haven't had a craft where nothing fails for a while now. What really got me suspicious now was a Duna lander with twelve 24-77s, where eight ( ! ) broke after just a short deorbit burn, landing assist burn to help the chutes, and attempt to hop a few hundred meters so in the span of being active for maybe a minute and a couple restarts (not sure if relevant). Age also shows ridiculous values like 40000 which is about the total mission time if this is in hours, and MTBF of 0.16 despite the fact that engines are supposed to only age while being active. Not sure if there's a problem with mod interactions, i do have a lot of other mods installed too, but don't really have the energy to figure this out at the moment so i'm just gonna uninstall it for now, but i wanted to let you know. Cheers for an otherwise great mod that enhances the atmosphere a lot.

Works for me, also for already existing crafts (or well the one i tried just now which did not work before). Thanks

I don't suppose there's a way to make a part a root part during active flight? I did select "control from here" for the probe core but that also did not help after undocking from the space station.

I have the same problem and do use SETIctt. One addition to this bug: I managed to launch a 3 man pod with 3 tourists with a OKTO2 probe core to my space station, only when trying to undock and fly back did i get this problem.

I am new to all of this and learning as i go, so thanks again, you are really, really helpful and this cleared things up yet again What i wanted to say was i need to think about what exactly the real part of the quaternion would stand for since that would be the roll information i currently store with the separate starboard vector. So basically, how is the craft rolled when the real part is zero since that would be the basis for everything and i need to know how to initialize it at the beginning of the simulation. I just don't have a (visual) grasp of it just yet. As for errors, there is so much stuff going on that i can easily just normalize each tick, it really makes no noticeable difference.

Thank you, that sounds all very helpful. Now i just need to think about how exactly to store the ship's orientation in a quaternion.

Science is a method and a way of thinking. By making observations, formulating hypotheses and testing those (by seeing if predictions from the hypothesis are true), it is the only tool to find out what reality is like. As for some of your secondary questions: What will it allow us to do in the future? That's impossible to say. It's hard enough predicting what we will find out in ten years, let alone in the distant future (at least that is what your question sounds like to me). I am convinced of one thing though, if humanity lives long enough it will discover many exciting and strange new things beyond our wildest dreams. Think something like quantum mechanics, or the theory of relativity, only one step further still. Are we using it properly? "We" as in humanity...not completely. You just have to look at all the misuse of science in the form of weapons. Really not a surprise if you look at humanity today though, maybe one day we will grow up and only use science for good. What is it's potential? Pure speculation, but science may one day fully explain our universe and maybe even beyond that. If we as humans are intelligent enough to understand it and get this far is another question.

I'm using quaternions. And since i'm still pretty new to programming (this is my first project bigger than "calculate some prime numbers"), low level coding does indeed sound scary. Besides, if performance were a problem i could just reorthonormalize only every 100 or 1000 rotations. The error shouldn't be noticeable then either, the only reason i am doing it every 10 rotations is because there is no noticeable difference in total simulation performance. And branching...well, again, as long as i don't notice a drop in performance it should be fine.

Just wanted to let you know, it works like a charm. I can easily afford to reorthonormalize with Gram-Schmidt every couple rotations (perks of having a very computation heavy simulation ^^), doing it every 10 rotations atm, where the floating point errors are obviously pretty much nonexistent.

@PB666: I don't have much to say about all this since i have yet to dig into this topic, but thank you so much for writing that out, that was really helpful. @K^2: Currently i have set the origin of the coordinate system to the starting point of the Sun. I don't see how this leads to problems though, i have a fixed frame of reference and calculate everything from there. That shouldn't be a problem, or is it?

@PB666: I can't follow you completely, but i plan to consider every major body for n-body physics for now. If the performance gets too low i can still change that, Pluto is a good example for a body that won't have a significant impact most of the time. Here's how i currently plan to deal with n-body gravity: I have only started to look into this (so many things to learn for a simulation like this...) so this is all subject to change, but so far Encke's method seems to be still relatively simple, yet pretty accurate. Basically you predict the unperturbed orbit which you can do exactly, and then introduce the perturbations that change the orbit. The error being proportional to the perturbations means it will also be small since perturbations are small in general so it sounds pretty neat so far.