Kersplattle

I have to say that I *HAVE* had issues, but am not for now. It may help that I have 16 GB RAM and the GTX780M GPU on a 27" iMac. I fixed most of my issues with removing stray kext files, the problems were mostly fan spin ups with 400% processor utilisation rather than out of memory and KSP crashes. Then again, if it *IS* a Unity 4 bug then it cannot easily be solved in KSP and likely enough the Unity developers are not focussed on v4 now. I'm prepared to be patient for U5, at least for now. Thanks for the attention on this thread from Salvager and others.

Well, not sure if i solved it fully but I think I was somehow being so accurate I kept colliding with the target capsule. Best explanation I can come up with. This last time I stopped 1.2 km away and jet-packed the stranded verbal (Moena) over. That worked, managed to get her home with about 1 unit of electricity left !

Not sure if this is just running out of time, but every time I get within about 0.5 km of the capsule that is my target it disappears and the contract is a failure. Why this is a problem is that by using F5/F9 I can change the elapsed time and I always get the same result. Each time I get a close intercept regardless of how long it takes it fails. Could this be that the contract actually has already expired but that isn't recognised until I get close enough to switch to the target ? Just FYI, this is the knarliest rescue I've yet seen. Target is a capsule not a Kerbal in a suit, and the orbit is at around 10 degrees off equatorial and is out between Mun and Minmus. Took an upgrade to go beyond 30 parts to be able to built a craft with enough fuel to reach.

Well, I just started a new career mode with the normal setting to gain some familiarity. So far so good, new features are pretty well integrated, and not too many bugs running the OSX version. Did get the bug where if the capsule hatch is slightly overlapped by a parachute (it an be anything but for me that's what it usually is) when you EVA on land Jeb explodes ! It does seem hard to gain science compared to the last version, very few science points on offering the contracts, I probably have to do a couple of limited sub-orbital missions to get science from local biomes. Or start making planes to get the local survey missions. Also wondering about the tourist and VIP contracts, they seem to say that these people need to be delivered to a location but no location specified. Could just be a language issue and later contracts will have a specific location; for these early contracts it is just a sub-orbital flight. Found the changing thrust more of a challenge for early flight, seems to force one to use Solid boosters for take off, but maybe that's realistic. Overall pretty solid and I like it. Though I have yet to get past the first layer or two of the tech tree and explore how that is now laid out.

The standard Apple wireless mouse supports left click as a left click and right click as right click, so to speak. It is sensitive to which side of the no button mouse you click on. Works well but can be a pain when you click near the middle and it decides you clicked right when you want left and vice versa ! The older non-wireless mice may not support this. Also, I posted earlier about 10.10.2 being more stable, which was a bit premature. However I have found that a Google app, Soundflower, the audio component of a screen capture app was very likely a proble. By removing that app and its associated .kext file I haven't had any kernel panics or crashes for over a week. So if you have that Google app you may want to try removing it if you are having stability issues. YMMV of course.

Just to note that I have had a lot of the same issues, only it is the Mac that gets borked. High CPU (Kernel Task at 350-400%). It's a late 2013 27" iMac with the Nvidia GTX 780M video card, the 3.4 Ghz i5 and 16 GB RAM. It seemed to be KSP that tipped it over but then it would reboot still in trouble. I have uninstalled Google Soundflower, a screen capture sound add-in and that seems (for now !) to have stabilized it significantly. No crashes for nearly a week nd it was daily before that, with multiple OS reloads. So just in case you have a program that adds a third party .kext file (Kernel Extension) to start up it can be the cause of a lot of problems. KSP and Unity don't seem to have kexts, so as I seem to have found, check for any third party (non-Apple, non-NVidia if you have an Nvidia card) and try uninstalling that app/program. Just FYI and YMMV, but it saved my KSP experience.

I use (or try to) KSP on OSX 10.10.2 and it is currently unusable. System crashes repeatedly with kernel run-away (hmm, sounds like kerbal runway) processes that is much worse in .90 in 10.10.2. I'm inclined to blame OSX because I can get it to do this a lot even without starting KSP, though it could be a specific kext file from Unity or KSP causing the problem. It's a fairly recent Mac too, late 2013 with 16 GB RAM and the updated GPU with 4 GB, when it works it handles pretty large craft without to much trouble, but the "when it works" part has become the problem just lately. Still looking for some sort of solution as I am reluctant to go out and get parallels and a copy of Windows, but that maybe the best answer. I suppose I could try a Linux desktop with Parallels and see how that performs. So I'd be interested in the same answer, by the sound of it either Win or Linux would work fine. Which Linux is recommended, vanilla Ubuntu ?

Just an update to an earlier post of mine, KSP on OSX *seems* to have become more stable with the latest update to OSX to 10.10.2. Haven't had to reload the OS for couple of weeks now which is an improvement. Worth trying if you are experiencing a lot of problems and are on 10.10.

Not sure if others are having this problem, but I have a late 2013 iMac 27" with 3.4 i5, 16 GB RAM, NVidia GTX 780M with 4096 RAM, and it runs Kerbal well, but the machine regularly has a runaway kernel process that takes up 300-400% of CPU making it useless for anything. Restart can rarely fix, but once it starts it typically continues on restart. Only fix I have found is a complete reinstall of OSX 10.10.1. Could this be Kerbal or Unity related as there's no a lot else non standard loaded. Kerbal crashed when trying to launch a largish but not huge ship yesterday, and after the system was restarted the kernel run-away is back. (That's "kernel run-away" not kerbal runway, just fwiw).

Just for info, I have exactly the same issue although I'm a little closer at 6100m, but after faffing around for ages trying to get over the highest piece of Gilly nearby on the hope that it would help, I simply can't get a scan done. As it is actually impossible I think it is debug time...

Need to modify that Australia option to either Australasia or Oceania. Us K1w1's don't like being lumped in with Aussies, and vice versa. So no vote from this user.

How about a craft management "system", so one could use sub-folders or some such method for managing all the various crafts one assembles for fun and profit. I often finish up with 50-100 craft and damned if I can remember which is which sometimes. Making weird variations is half the fun, or at least to me it is. And I like the ability to recover discarded stages if setup to be recoverable with chutes etc. It's OK with stages that reach orbit (or nearly so) as one can switch to them and bring them down, but not always.

I don't think its the RAM or Graphics, I have an iMac with a 3.4 GHz i5 (QC), 16 GB RAM, and an NVIDIA GeForce GTX780M with 4096 MB, and I get these freezes a lot. Using Yosemite 10.10. Happens anywhere. Recently happening a lot when building rockets; also when parachuting to a landing. Sometimes I get a kernel process spun off taking all the CPU, sometimes just the "SBBoD". Only ever happens when KSP is up. Making it hard to play as I rarely get more than a couple of hours without it happening. Used to happen in 24.2, but much more frequent in 25. There is something in KSP doing this, not sure what but it would be great to have it fixed.

I confess to not being able to use Hohmann transfers and I make all my interplanetary missions by going into solar orbit and then burning to get an intercept from there. Uses a lot more fuel so I compensate by building bigger ships. I do get frustrated at the slow acceleration when using efficient but underpowered ships with nuke engines to go to distant destinations like Jool. I also use F5/F9 to get the right descent to a planet or moon and still consider the mission a success ! And this one is not really a KSP sin, but my answer to most launch problems is...MORE BOOSTERS ! Love them "trash cans of boom". Though every now and then they become "trash cans of doom" !

Started afresh in 23.5 in career mode, went through the tree OK completing it with a trip to Duna. The new heavy engines and fuel tanks make building a heavy lifted sooo much easier. Then it was off to Laythe and Vall, two missions, same basic lander and tug stage. Because there's no aero braking at Vall it is a lot more expensive on fuel and I should have swapped the aero spikes (chosen with Laythe in mind) for nukes, but both missions completed successfully. Next up, probably Moho with the nuke equipped lander, and I'm trying to build a successful Eve lander that can take off again. Not easy, using probes the size gets prohibitive very quickly. This is all completely stock. Haven't done the asteroid bit yet, maybe later, seems a bit "meh" to me.

No comments about the ship in particular, but one solution I've used is to have 4 engines, 2 x Nuclear and 2 x Aerospikes, with action group activate/deactivate. The spikes are better in atmosphere, more efficient and with 3 x thrust. Aero braking on Duna, aim for a low of, say, 17,500-18,000 for the first pass, actually 16,500-17,000 is OK and brakes better, but YMMV, test your crafts reaction. Then you need almost no fuel to land on Duna if you target the low lying areas to land in so you have enough time to slow down. I fit 4 x drogue chutes on a stage to trigger early (10-12K), and depending in the mass of the ship, 4-8 normal chutes deployed once you're travelling slow enough. The rogues fully deploy at about 3K up, the standards at 500 m from the ground. You may need engines to decelerate to deploy (or may not) and will probably need them to slow the last 100-150 m to the ground. Use the internal camera © to watch the radar altimeter to get a good feel for the ground height. I can usually get back and soft land (re packing the chutes is a must !) with either my heavy one piece lander or the light lander plus docking stage. Last time I damaged the docking interplanetary stage and actually made it back fine with just the lander and the fuel left after landing. I'll post pics if I can find them.

Feeling pretty please as I managed to get a lander onto Eeloo and return to a soft landing on Kerbin, admittedly not at the space port. This was a significant challenge in terms of fuel capacity as I like to run stock and sans Mechjeb; so it's all hand crafted piloting with the "add manoeuvre" flight planner. I hope to post some pics when I'm back on the gaming machine. So much fuel use to burn into orbit when I arrived out there, and over 3000 dV to get back into Kerbin orbit. I really must plan my intercepts better. The design uses my heavy lifter asparagus core which is only 4 x symmetry and discards all 4 in a layer at a time, so it's not super efficient, and has bulk SFB's to get it started, plus an interplanetary stage docked to a smaller lander. It was only when lifting off from Eeloo and trying to make rendezvous with the interplanetary stage to refuel that I discovered a design flaw, no SAS on that stage so sooo hard to get it stable enough to dock back onto. Judicious use of 5 x warp to stop it spinning was the solution and eventually we docked, with 0.5 units of mono propellant left in the probe controlled stage. Then fuel transfer to suck it nearly dry, discard most of fuel tanks (just keeping the central one with the last of its fuel, and home using the lander engines, throwing off the last tank half through the deceleration burn to intersect with Kerbin. This does take a refuelling in LKO of the remaining core of the asparagus lifter to give me enough fuel to get into solar orbit before pushing out to Eeloo, but it sure is a long mission.

Herr Doktor Kersplattle agrees with this sentiment, anyway he hasn't stopped. I'd rather work out a better controlled way to get a large ship to make decent gravity turn and not go half out of control. Maybe a new Mac with a better graphics card is needed. Serious lag problem.

Anyone else having problems with the nuclear motors dropping off when in flight ? Usually while in deep space and trying to turn the ship around, one or more of the motors just fall off. Nothing to collide with, and they have about 4 struts attaching each to the relevant fuel tank. Ruined several perfectly good missions before making rendezvous. Thanks

I originally constructed a 3 man craft so I could rescue stranded kerbanauts ! I'm an empirical constructor although I will get around to doing the math at some point. Strictly stock construction, attached is a pic and a craft file. Biggest problem is keeping frame rates at a point where the altitude meter doesn't go haywire, and stopping the craft disintegrating on ascent with a failure between fuel tanks. Works fine otherwise, has fuel for Minimus and Mun in one mission. Can be a cow to control with the lower stages in place though. I really should rebuild to refine, but as it works I haven't yet done that. Currently trying to launch enough fuel with a one man lander to do a 4 x landing, Min/Mun/Min/Mun and get home, three is easy enough, 4 has been a bit marginal on fuel. Can't wait for 0.17 and some further challenges.

How about when at 10 metres up and descending nicely to Mun, in your haste to make a small correction in heading you hit G instead of F and your gear smoothly retracts just as you hit....

If you really want a short bit of coaching the tutorials are good. But here\'s my best go at a short description of how I approach it. First build a lander. For first attempts make it small and light, that way it\'s easier to change velocity though they can be very sensitive on the controls. Try the smallest cockpit for one kerbalnaut, add one or more parachutes, then a decoupler, then an ASAS and say one large 'small' fuel tank and the smallest engine. For refinement add an RCS tank and say 3 thruster blocks. Add 3 of the larger legs (heavier but much more resilient, the small can be used but tend to break if you have any transverse velocity). That\'s the lander; might be a bit short of fuel to return, you may try a second small tank, experiment. Next you need the motive stages, and this is often a case of trying a certain configuration and modifying. For first goes keep it simple. I\'d go for two stages here for simplicity, with one to launch and 1 to push you into orbit and to Mun. The second stage is, say a couple of BIG tanks and a big engine, attach with a small decoupler (I reckon these are more stable, but YMMV). When experimenting if you need stability add fins and struts for support, just be sensible. I\'d add fins for the early part of the descent too. First stage, try the same as the second plus, say, 3-4 of the larger solid fuel boosters, plus fins. Attach with small or large radial decouples, I like the small as they seem to detach better, but that may be just me. Pay careful attention to your staging; have the first stage and the solid fuels in the bottom, then a decoupler stage to drop the solid fuels, then group the first stage decoupler with the next stage so you just drop the stage and keep going. You may need more push, you could try a small third stage above the second, but experiment, you can often add solid fuels for just a bit more but try not to get too complex. As you go up, start to head North at any time from 8,000 metres up, might be best to wait until 12K or 16K, but I\'d say definitely by 30-35K where you can be over by 45 degrees. This reduces the 'gravity drag' and reduces fuel burn. Wait until you have an apoapsis of, say 80K (suit yourself on this, can be as low as 70 or as high as 100-120 to 240 if you have plenty of fuel). Stop burning and coast until you are at the apoapsis ( 10 or so secs out to give yourself a little margin) and be ready at close to a horizontal burn, a little up and adjust down as you go, try to keep close to the apoapsis until you get a periapsis and burn until that\'s above the atmosphere. Your now in a stable orbit around the equator or close to it. To get to Mun, slow way is to go to your periapsis and burn towards the round green circle on the control ball or whatever you call it. That will extend your orbit on the opposite side of Kerbin, push it out until it intersects the Mun\'s orbit, and warp time until you pick up an intersection with Mun. Faster way is to go into the rocket view, orient yourself until you are pointing at the green circle in your orbit and change the view to look forward at the Kerbin horizon. Wait until Mun rises and immediately burn at full thrust. You should intersect with Mun the moment you orbit crosses Mun\'s. I\'d then warp until you are over by Mun and right next to the perhaps is with Mun, then decelerate until you get an orbit. Adjust your orbit by burning retrograde as appropriate, into a reasonably circular orbit. Try to get the orbit so it crosses one of the large craters, they\'re pretty smooth to land in. Next to get to the surface, the\'s lots of ways. I like to choose a spot, then wait until I\'m say 10-20 degrees short of being over the top of it, then decelerate hard until your orbit intersects Mun, and keep burning until your intersection is about where you want to go. Remember that it may move further as you brake when you descend depending on how much lateral movement you retain. Now I often switch to the spacecraft view and warp until, say, 50K or so, burn to reduce speed to under, say, 50-60 m/s, then switch off and drop to say, 20, repeat, etc. you can have fewer longer burns, or if confident just one final burn, experiment ! At say 4-5 K up you want to be pretty much dropping vertically, I like to keep the speed down here to 50 max but there\'s no hard and fast rules, just be sure to have enough time to slow down or, kersplattle ! Lower the landing legs at, say, 4K, and guide it down. If you keep aimed at the green circle with the cross you will remove all lateral velocity but it is hard as your speed drops to below 5-6 m/s to stay stable, as above, work on it. You can use RCS if you have it, it isn\'t essential. At the end try to get to around 2 m/s for the last 20 metres, it\'s a judgement call and you\'ll get better as you practice In my install the surface of Mun is at A minimum of around 720-750 metres, often around 850, and that\'s in the craters. Can be up to 2K outside or more, hard to be sure but watch the terrain change. That\'s it, you should be down. Next crank up the throttle and exit if you have the fuel. Aim at one of the pink items on the trackball to get an equatorial orbit. Get into low Mun orbit, burn to accelerate backwards, exit Mun orbit then got to your Kerbin apoapsis and burn retrograde until your orbit intersect Kerbin. Go home. Phew, that\'s much longer than I thought. Sorry admins if it is too long.

Well, could be a little unsymmetrical at the bottom stage, I\'ve found that it helps with larger designs to avoid instability at lift off to use RCS until you get to,say, 40 m/s. That does tend to imply that the fins are needed for directional stability and that the thrust is not quite as symmetric as I would like. I can build smaller craft, I want to lift more into orbit though. Might try using more large parts and reduce the staging. There\'s a fairly fine line I\'ve found between a reasonable stable craft and something with just a bit more power. Can be frustrating trying to get the balance just right. I have a 0.15 craft that did about 20 Mun and/or Minimus landings quite easily, and want to get someone larger to enable more flexibility. Doing a Mun then Minimus, which seems a harder mission than Minimus then Mun. Thanks again though for the feedback.

It doesn\'t wobble excessively, but it does a bit. The RCS tanks are just to give me enough RCS fuel to start lift off with RCS on, so many to remain symmetrical, may not be necessary with this version of the craft. An earlier version had a smaller lander stage at the top, less height. It might be that it does get unstable at the top as that version was more stable, but had a smaller first stage. It enabled a Minimus landing, take-off, traverse to Mun, land, and take off with the final stage and make it home, but it was quite marginal fuelwise on the Mun landing, hence the attempt to lift a further few tanks into orbit. I\'ll experiment a bit with the suggestions, thanks. Bit late now though, more tomorrow.

I have a biggish rocket with 6 large engines as a first stage, and about 15,000 metres up or more the rocket just starts to tilt and gyrate out of control. Having RCS on makes no difference, I have 3 ASAS modules, lots of fins (which maybe why it\'s stable while the atmosphere is 'thick'). Once it starts to tumble nothing stops it, basically it\'s abort or end flight. Any ideas why it might suddenly go astray ? The craft seems pretty symmetrical. It has a second, third, and lander stages above all using small (1 m) parts, and the connection is 3 large decouplers connected to the second stage. The second and above stages have all flown successfully before in 0.15 and 0.16 on other first stages. This is on the unmodded 0.16 Mac version and this rocket/model has serious lag issues at take-off.