Starman4308

Thanks for the response: I dived into installing Linux, assuming my year of doing command-line work on an iMac and a Linux supercompute cluster prepared me for it. I probably should have done at least a bit of research, so I wouldn't be surprised by things like "hey, what's this apt-get command?": a Linux distribution is more than its command line. I might switch to a different Linux distribution eventually, but for now, I'm content with having it work. On that specific bug: I'm 95% sure that by the end, I was only having issues in having Bumblebee force anti-aliasing. Just be sure you're consistent with using primusrun; optirun won't use the primus library which further optimizes what Bumblebee is doing. I'm 95% sure I have the latest NVIDIA driver for Ubuntu (331.38); that's easy enough to get from the command line (just sudo apt-get install nvidia-331 nvidia-settings); the nvidia-prime package is what I use to set GPU vs. onboard chip now. I think there might be more up-to-date drivers on the NVIDIA website, but after seven hours of fumbling with this, I'm content just to have it working. On the shell script to launch it: sure, those Steam libraries might have been why KSP was balking at the 38 mods I foolishly decided to simultaneously install right off the bat. Otherwise it's back to "uninstall everything and install one-at-a-time". On more amusing notes, I spent 15-20 minutes searching places like /etc/ and /usr/share for my JDK, which happened to be in my home directory, and I'll probably alias the shell script to something like "moar_struts" or "rapid_unplanned_disassembly".

Okay, to help eliminate confusion, can you clearly separate what you reentry profile is from what speed/altitude you pop chutes? Say something like "My reentries were 80x40 km, and I popped chutes at 7 km, with speeds ~250 m/s"? Also, Toshogu, are those typos when you say "7000k and 3000k deployments"? A 7000m pre-deployment (into drogue mode) is sensible: a 7000 km pre-deployment means you're popping chutes in high orbit.

I'm wondering if that might not be a setting. I know the past few times, I had my chutes as the last stage, and I had to press spacebar a second time to arm them.

It's because numerous mod authors are fed up trying to provide support for Windows x64 users, where the bugs stem back to Windows x64 instead of their mods. And no, there's no setting: you could eliminate the check from the source code and recompile, but they're still not going to help you with any bugs found on Windows x64.

Well, I can report that, among 37 other candidate mods, Trajectories might be responsible for exploding rockets on my very recent Linux 64-bit installation. Useful feedback to follow if I narrow it down to Trajectories (unlikely).

In terms of programming it: yes. I don't know how to do it, but a simple conical part shouldn't be too hard to do. There are good reasons why heatshields aren't pointy, though.

You have to open up the Trajectories window in map view and play with reentry profile: by default, it calculates it with your craft pointing straight prograde all the way through.

It's strange that they would burn up at just ~200 m/s though: that's subsonic velocity, which is in the realm of sanity for when to pop chutes. There's also a max temperature multiplier of 0.25: the parachutes burn up a lot more easily than their casings. Are you sure you're only at 200 m/s surface velocity? Supersonic velocities (~330 m/s) are the destroyer of parachutes.

Yes. And apparently the forums don't like laconic, <10 character replies.

I did mostly play with a blatantly reduced heat transfer coefficient (10-11) using the 6.4x RSS config so that I didn't have to bother with the RO heatshields: whether or not I should've set it higher is up to debate. I did usually tend towards going shallower than 50% atmosphere in any event. As to apoapsis: maybe I'm wrong, but it seems sometimes you can be gentler with a high apoapsis if you do it right and "bounce" off your periapsis, killing even more velocity in upper atmosphere. In any case, just remember: do your level best to bleed off as much speed in the thin upper atmosphere as possible, because the lower atmosphere will kill you if you are still moving at anything remotely resembling orbital velocity. There is simply that much more air to transfer thermal energy and bleed away the ablative heat shielding. To clear up one point: what would burn your hand faster, near-vacuum at 1000 degrees Celsius, or a pot of boiling water at 100 degrees Celsius? It isn't the near-vacuum, because there is so little matter to transfer heat there.

EDIT: For my definition of fixed, it's fixed for me now. I uninstalled Bumblebee, installed nvidia-prime, logged in and out, got a nice fancy options window, and forcing AA seems to work from there. I don't much care that it's keeping the GPU on 24/7, because this laptop has pretty much become a fixture of my desk due to being too big: it was a relic of my undergrad days when I'd be shipping it back and forth from home. Has anybody here been successful with using NVIDIA Optimus technology with the Linux x64 build? I decided to add Ubuntu to my Windows 8.1 box (finished as of this morning), and while I am 99% sure I finally got it all patched together using Bumblebee and Primus, the terrain still seems very rough, which as far as I can tell, is because there's no anti-aliasing. I've even gotten to NVIDIA settings (using optirun nvidia-settings -c :8, because regular nvidia-settings breaks Bumblebee) and set 4xSS, 2xMS, and 4x anisotropic filtering. The only two possibilities I can think of are that those over-ride settings aren't getting properly read, or I'm not properly setting OpenGL. Steam launch options: %LD_PRELOAD="libpthread.so.0 libGL.so.1" __GL_THREADED_OPTIMIZATIONS=1 ./KSP.x86_64%

And there's your problem: you are coming in way too steep. In my experience, setting periapsis halfway into atmosphere is a decent guideline, and I often use even shallower reentries. You kind of want to bounce: you want to shed as much velocity as possible in relatively thin upper atmosphere. Trajectories is a pretty decent mod for figuring out where you will reenter, though it doesn't work with stock lift (only FAR/NEAR), and I'm not sure it models loss of ablative material. The second caveat should not be a big deal, because if you're doing reentry right, by the time you start losing buckets of ablative material, you've already shed most of your speed.

You can always try a lifting re-entry: the goal is to have an asymmetric pod, and roll it so that the heavy side of your reentry pod is pointed upwards. Because your CoM is out of line with the drag vector (intensified by hypersonic lift generated by the heatshield via strange means), your pod will wind up with the heatshield pointed a bit down and the top of your capsule a bit up. Because of that, FAR generates a bit of lift which keeps you higher in orbit longer. Symmetric pods have a hard time because reaction wheels just don't give enough torque to move the pod out-of-line with the retrograde vector. You can also abuse this to shorten your return, by rolling the other way and causing lift the other way. You may also need RO-rated heatshields, I don't know.

I actually did that once: built a drone with a refueling port, a claw, and an LV-N, and went around de-orbiting all sorts of junk. Less of an issue now that I'm using RSS: most of my stages wind up never getting to orbit, and my current crop of orbital rockets tend to use the last ascent stage for all their orbital maneuvers and re-entry.

Here's my 1-way science probe: relied on NEAR, Procedural Fairings, and I used RealChute for stage recovery. Landed this one with 300 m/s remaining, even with an accidental dip into upper atmosphere after ascent and a Munar transfer from a non-circularized orbit.

I don't know the numbers, but I do know the theory of what's going on. There's a closed loop: the plants take in carbon dioxide and produce an equivalent quantity of oxygen and fixed carbon. The colonists and various decomposers will consume the fixed carbon and oxygen and produce carbon dioxide. At excess oxygen, CO2 is depleted and the plants stop growing. At excess CO2, the plants love it, and the colonists die, leaving the hardier decomposers to survive. If well-planned, there will be a sufficient buffer of CO2 and O2 such that, between planting and consuming crops, neither O2 nor CO2 is dangerously excessive or depleted. This could be assisted by using staggered plantings, such that only a fraction of your crops are just being planted or being harvested at any given time. Granted that this is Mars One, in the absurd off-chance it actually gets to Mars, it probably will not be so planned, and will quite possibly be the cause of their eventual demise. Mars One was never a good idea.

When landing on Duna, parachutes work great thanks to the presence of atmosphere. Just remember: if you're using Deadly Reentry, be sure to pop your chutes late: do it too early, and reentry will burn your parachutes off, just like on Kerbin. Test things out home at Kerbin, where it is quick, cheap, and easy.

Try deleting and re-installing Trajectories. That worked for me. Also, I am 95% sure loss of ablative heat shielding is part of why I always land short of Trajectories's prediction: the heat shields are almost solid ablative material, with very little structural mass to them.

That depends on velocity and number of intakes: there is no one constant altitude to cut your jets. If you use Mechjeb, in the Utilities section there's an option to have MJ throttle your jet engines to avoid flameout for you, and you can edit a window to show you intake air, amount of intake air needed, and amount of intake air with all intakes open (MJ also has something to open/close intakes for you to minimize drag).

By the way, how much effort would it take to fudge the rewards for contracts outside Kerbin specifically? I'm playing with the 6.4x RSS config, which means that there is a very, very sharp increase in the difficulty of extra-atmospheric contracts, but anything atmospheric is exactly as easy as stock (probably moreso with NEAR removing the souposphere). It's not terribly fun when I either get bankrupted by extra-atmospheric contracts (low fund setting) or roll in money from atmospheric contracts (high fund setting).

Congratulations! That must've been a very tricky rendezvous. Must've been a bit difficult to manage reentry without frying your engine as well: usually, for me, I ditch anything except the command pod, parachute, maybe science payload, and the heatshield before reentry.

So, if I'm understanding it right, at the tail end of your reentry, temperature is lower but thermal energy flux (heat flow) is greater because there is that much more air slamming into your heat shield, even if it's not as hot, and because ablation rate is (presumably) proportional to thermal flux instead of raw temperature, that's when you lose most of it?

RealChute should not be doing that, because RealChute just adds some parachutes, and in theory, functionality to existing parachutes. Are you referring to Deadly Reentry? In that case: make sure your landing gear is tucked away behind a heatshield: I will often use a procedural heatshield from Procedural Parts which is just a bit wider than the rest of my craft. Anything which sticks out more than the tiniest bit from your heatshield is liable to fry if you're using Deadly Reentry.

So, quick question: what factors influence how quickly ablative shielding wears away? I've noticed on my past two re-entries that I use 20% of my ablative shielding is used for 90% of the descent, with the rest of it consumed in maybe 30 seconds once I hit ~25 km (using the 6.4x config of RSS), despite the fact that velocity and temperature are both declining by that point.

Got rolling on a new career mode game using the 6.4x config of RSS, NEAR, DRE, and TAC Life Support (all of which are new to me). Launch vehicles have grown substantially larger, and every kilogram of payload is now a torturous decision, as payload to the Mun has gone down to ~0.3% of total rocket mass. I gained ~700 m/s dV on my latest rocket by switching out an LV-T45 for an LV-909 on lunar transfer stage, though I re-learned the lesson of "TWR is important always": I wasted a bit of that dV because I wasn't able to circularize without a short excursion back into upper atmosphere. A short, fiery excursion back into the upper atmosphere. Nevertheless, Luna 9 is in transit to the Mun has juuust enough fuel to stick a landing... so long as I avoid wasting the gobs of fuel I'm used to on descents.