vossiewulf

I used autostruts in one game for a period of time, and had horrible problems with space stations going kraken, especially when releasing large orbital-built ships. I strongly suggest anyone who builds big to stay away from autostruts and stick with KJR.

Just one problem, autostruts have been known for as long as they've been around to be highly krakenating. It is much better to use the Kerbal Joint Reinforcement mod and never have to think about wiggly rockets ever again. And easiest way to get that installed is to use CKAN and select KJR from the list of available mods and hit the install button.

Duke, I've noticed some landing legs are more prone to walking around than others. My current favorites are legs in the Modular Rocket Systems mod (can get it from CKAN, current for 1.81 at least). The legs have good shock absorbers and that seems to let the vehicles wiggle around a bit without moving their feet. OTOH, all of the SpaceX-type legs I've tried love to take long walks across the surface. Also make sure you're set to the highest level of terrain detail, that helps stabilize things too.

For local launches (Mun/Minmus), I think the windows are much too small for multiple launches, so I launch, warp forward to next window, then do the next launch and anything non-local gets launched to the target from orbit with reasonable separations (hours apart, I don't want to be madly switching back and forth between vehicles as they all hit the target SOI close together). If you're willing to make a change, for simplicity's sake just add one minute padding to launchpad warps. That's plenty to do one last scan before launch, and if you're warping from the launchpad you're saying you're ready to launch so you shouldn't need much padding. I think the padding for Space Center warps should be a bit more since you may want to hit the tracking station or mission control first. In orbit, where you know the window, I think your current logic works fine.

Hebarusan, is that behavior you're describing from the Space Center, or the launch pad? I tend to use it on the launch pad, i.e., launch and then warp to window then take off. In the latter case of using it on the launchpad, the vast majority of the time it warps right to zero on the launch window, and in a small percentage of cases you'll see it go to zero then suddenly there are a few more seconds back on the clock (up to 5 max) that then tick down again. I've just done a whole series of these warps from 9 hours ahead of time down to 30 minutes and I see the same thing every time, it doesn't leave you time to nail the takeoff window. Keep in mind though that I also use Astrogator constantly and it's one of my required mods, so just trying to make it better here (at least if what I'm seeing is not intended), and thanks for making it and supporting it.

Either use Mechjeb's Maneuver Planner> Advanced transfer to another planet function (just target your intended planet and tell it to create a maneuver node) or install Astrogator, both will identify launch windows and plot exact courses for you. I think Mechjeb is a bit more accurate for transfers, but Astrogator has a window that lists off all the upcoming transfer windows in one place, making it easy to figure out what missions you have to launch and in what order, so the compliment each other.

A quick question, why does the USI-LS habitation-monitoring window almost always show me that my space stations are out of EC when they're nowhere close to being out of EC?

SSTU, Blue Dog Design Bureau, everything Wild Blue Industries, Astrogator, Waypoint Manager, Trajectories, Atmospheric Autopilot, Kerbal Engineer, everything made by Nertea (Stockalike Station Parts Redux, Near Future everything), Contares everything, DMagic everything (ScanSat in particular), Distant Object Enhancement, Planetshine, EVE, Scatterer, Kerbal Foundries, SpaceY, B9 Aerospace, and of course Mechjeb.

It's more like finally using the GPU, instead of doing nearly everything with the CPU like KSP1. I haven't seen anything so far in the released videos that says to me that KSP2 will push any recent video cards though. So the good news out of that proper rebalancing of processing responsibilities is that some people with older systems may be able to still run it well with just a graphics card upgrade, and even then they could get something a couple years old and inexpensive and it will still probably be plenty- no need to rush out and get an RTX2080, KSP2 wouldn't even make it break a sweat.

What the heck, worst that happens is Sarbian doesn't have time. This is what I see from the landing process once you pick a location on the ground and click on Land at Target: He warps to the plane change without first orienting himself for the burn. Therefore it can miss its burn time if the vehicle is large and slow to turn. He warps to his first burn where he positions the landing point exactly on target. Again, he doesn't orient himself first so we have the same problem of missed burn times. He warps to his burn where he brings velocity to <1m/s to begin his drop straight down landing segment. Again, he fails to orient himself first and there's no way to make up for this burn being late- a slow-turning vehicle is probably going to crash. Then, for some odd reason, he first orients himself for the final suicide burn before warping to said burn. That makes this one the only accurate one in the group of four burns. The workaround for me is to disable auto-warp, because in burns 2 and 3, he has himself set to basic SAS retrograde. I then selectively warp, stopping well before his burn points to give the vehicle time to get reoriented before the burn. If I do this semi-manual process, I can get him to land within 400m or so of the target, even with large vehicles. Finally, on burn 2 he positions his landing target directly on the intended target. This will always make him land several hundred meters short, because of burn 3. Burn 3 brings him to a dead stop 1500m or so up, which means he drops straight down well short of his target. If you set the target in burn 2 to a bit past the intended target, then burn 3 would bring it right onto the target and you'd have a very accurate landing autopilot.

Are you willing to look into how the landing guidance works? If so I can provide some detailed feedback based on recent use. It has some basic problems in how it works that could be fixed that cause many types of vehicles to miss landing targets by significant amounts.

Carnac the Magnificent says... only if they don't default to polar launches

That'd be a first for a top-class title. In other words, complete repurposed bovine waste. Game companies always tell some whoppers but that's a whopper that wouldn't fit in a long-bed truck. Besides the general increase in complexity, it won't ship anywhere near optimized, nothing does, there's always a time crunch at the end and only the very worst perf testing results will be addressed prior to release. The way the industry works, no game is nearly as good or as performant as it could be when it's released. This. No one ever won an award for a sequel that could run on hardware designed for the prequel. You win awards for more amazing glitzy and spiffy graphics and lots of new features with considerable depth. Just like with every new generation of games, unfortunately some people are left out in the cold with hardware that simply can't keep up. The good news in this case is KSP1 is still KSP1 and lots of fun to play with a pretty limitless set of achievable goals.

Thanks. This wasn't aerocapture, it was an attempted landing from a 300km orbit. A "heavy" setting SSTU heat shield would probably do ok from a 65km orbit, but I don't know what to do about the tumbling- the vehicles was doing fine (except for running out of ablator) when it just tumbled end over end and that was all kinds of explody. I've recovered much taller (relatively speaking) vehicles on Kerbin with no problem, I didn't even consider this one as having a risk of tumbling. Next one gets serious reaction wheel overkill I guess.

What is the aerobraking/parachute protocol for Eve in JNSQ? I thought I was good setting my periapsis at 45k, but one I didn't have enough ablator, and two the craft tumbled at 35k even though it should have had plenty of reaction wheel torque to maintain its retrograde attitude. I was entering from a 300k orbit at about 5500m/s, which is fast but no faster than returning from Minmus to Kerbin. Also, if I haven't mentioned in a while, thanks again for all this work, to me this is the right scale to make things tricky without going to the self-flagellation lengths of RO.

Keep in mind that when I say three launches I mean pretty big launchers. Mostly Near Future Technologies and Station Parts Redux, but lots of others- Wild Blue Industries, SSTU, Station Science, various MKS and USI-LS parts, Global Construction.

This is a PXL stations parts and octo-girder-based extraplanetary launchpads construction station, only the docking ports and the workshops aren't NFT. I wish I could do the PXL parts more justice with a much bigger station, but I've been down that road and things eventually grind to a halt even with a very fast processor, no fault of NFT. This station was only three launches, nice and efficient. But anyway, just a screenshot of someone happily using your work Nertea, thanks. This one was very pretty, but became unusably slow... sigh. Didn't help that I gave it like 50 docking ports.

Their eyes are farther apart, making the eyeballs wrap around the side of their heads more. I agree it doesn't look better.

You know folks, it's a game, and they can do whatever the hell they want to with it and their only responsibility to their employers is to try hard to make a game that sells as well as possible. They don't owe you, me, or anyone else anything but that, and the extent of your control over the output consists of being willing or unwilling to pay for said output. Anyone taking this seriously really needs to go do something else until basic perspective is regained.

Thanks Hebarusan. I'm not quite sure it's doing that, but I will check now that I know what it's supposed to do.

Yes, it works fine. Question, when fast-forwarding using Astrogator to a launch window, sometimes the time warp stops 3 minutes or so before the window. Other times it stops 10 seconds or less from the launch time, and other times it fast-forwards all the way to zero so you end up missing the launch time by a few seconds. I've yet to be able to see what the pattern is, can anyone explain?

The chassis has built-in power, a 3U is about 6k EC. Plenty to run that thing for hours, and as noted, all I have to do is add a crew cabin (no crew) and it starts working so it should work without said cabin.

Is the auxiliary navigator probe core supposed to work on its own? Because the below doesn't work, it's DOA on launch and you have no control. However, if I add a crew cab and launch it without crew, it starts working.

Mechjeb>Landing Guidance. There is Pick Location on Map, there is a drop-down for standard landing locations, and there is numerical entry to enter coordinates if you want.

Oh I did do an orbital survey, I just stupidly didn't realize Rock was an official resource. The way Angel phrased it to me (slow, high in EC) made me think it was a fall-back converter that worked everywhere, and in return it was slow and used lots of EC. It was a dumb assumption, I didn't even look at the map overlay for Rock until you pointed that out to me