Troberg

Done it. The beast was just above 8000 tons on the launchpad... Six landers, seven orbiting science satellites. Just remember to put the mothership in a very high orbit around Jool, say halfway out among the moons. I didn't, and wasted lots of fuel getting to the moons. Won't do it again. The ascent was a nice slideshow...

Making a single launch mission that managed to deploy six satellites and six landers to the Jool system. I deployed one satellite around each moon and one around Jool, then landed on each of the moons, and used the sixth lander to crash into Jool, gathering science on the way down. All satellites and landers had a full set of science equipment a and everything needed to power it. The beast was over 7000 tons on the launch pad...

An integrated mod manager, making adding and removing mods simpler. Sure, it's not that hard, but the mod manager could do additional stuff, such as checking for compatibility with other mods and the current KSP version. Just specify a mod format (say, a zip with the files, as well as a manifest file (which tells which file goes where, compatibility info, author info, link to forum thread, link to latest version (heck, perhaps the mod manager could do the download/update as well) and so on). Then, the mod manager handles everything. That would make it so much less a pain in the exhaust nozzle to upgrade to a new version of KSP. Now, if you have a bunch of mods, digging around to check for compatibility with the new version is a lot of work.

I don't think so. I have my base as well, which I've deployed on Eve. That base is based on Hooligan Labs balloons. Basically, it's an airship base, which can drop off a smaller ground base, a bunch of rovers and a bunch of small airship rovers. These have been way out of 2.5 km of each other.

I didn't notice any difference between the unmanned test launches and fully crewed launch of my 164 kerbal permanent Eve base. That should have made a difference...

Just put a control pod or cockpit on the plane, and that won't happen. I often launch multi-part missions like that. My current mission is a mission to Jool, with 6 unmanned probes which are intended to land on the five moons, with the sixth probe plunging into Jool, collecting data as it goes down. The drive stage is left as a science satellite in Jool orbit. Nothing docked in space, it was all launched in one launch. Works nicely, done two landings so far.

I agree, MechJeb is not cheating. People play the game differently, I, for example, like the engineering aspects of the game, but don't care much for the actual manual flying. MechJeb allows me to play like that. I build great projects, huge bases, amazing interplanetary missions and so on, but I'm perfectly fine with MechJeb doing the flying for me.

Ah, then I'll scrap that plan for the moment (or do it anyway, because it's cool to shotgun a planet with probes). Or, maybe wait. I suspect biomes will appear in future versions.

Strange, when I had just landed on Ike, I saw a small shadow, about the size of a Kerbal head, zip past on the ground. At first, I thought it was the de-orbit stage, but it was nowhere near the right place for that. Possible easter egg or graphics bug?

One additional note: If you are doing an Eve mission, consider the possibility of making really minimal landing probes. I suspect you could do them at around 1.5 tons, science and all. They really just need chutes and enough engine/fuel to push them out of a low orbit. Then, just attach a bunch of them to your interplanetary drive stage, and when you arrive, make sure you get an inclined orbit (preferably polar). Then, it's just a matter of dropping off the probes on various places. You could easily get 6 or 8 such probes on your ship.

With that much of the tree unlocked, Duna and Eve should be a walk in the park if you make an unmanned mission. What I did was to make a rocket carrying two small unmanned probes on radial decouplers. Got into Duna orbit. Decoupled one probe and landed it on Duna (go in shallow for optimum aerobrake, use chutes, land on low terrain and you'll not need to spend much fuel). Decoupled the other and went to Ike, where I landed. Now, I had a satellite around Duna, a probe on the surface of Duna and a probe on the surface of Ike, all with a full set of science instruments. Make sure you remember to do science as you travel as well, then do science where they landed. I think my total payload for that mission was around 20 tons, lifter and interplanetary drive stage excluded. Not that much to launch. I think I'll tweak the probes a bit before Eve, making them rovers instead. That'll give me even more science.

Firecracker. The next rocket was Lawndart. Then came Auger, and after that I started naming them Rocket 1, Lifter 1, Skybase 1 and so on, with number increments for new versions.

I did a Mun landing (with return), a Minmus landing (with return) then sent an unmanned one way mission to Duna with three parts (one orbiting satellite, two small landing probes, for Duna and Ike). Not quite maxed out yet, but I plan to repeat the Duna mission on Eve and Gilly, and it will more than max out. Should be easier than Duna. The Duna mission alone brought me about 4000 science (just make sure you have all sensors and plenty of power).

Tried both. Also, tried to not limit to terminal velocity (that's the reason for the 2000 ton lifter...), the idea being that more engine thrust would provide more powerful steering to gimbaling engines. The thing is, as long as it's vertical, it's fine. But, as soon as it tips over, even if just a little, it will continue to tip over into a tumble. (I've actually almost made orbit once, despite doing two full tumbles on the way up, but I didn't manage to get it into a stable heading for the circulization burn before the orbit declined...) For really heavy vessel, it would be really nice if it was possible to use ordinary engines for steering, instead of RCS. Just put a bunch of sideways poodles on the ship (or, if you are Whackjob level insane, mainsails ), and have the RCS system control them. Perhaps some talanted modder out there who feels inspired?

My rocket is very tail-heavy, and has plenty of gimbaled engine power, all at the bottom. Also have plenty of RCS and a few fins. Depending on lifter configuration, it's a 1000-2000 ton vessel, with a 150 ton payload. Does it matter where the "control from" part is located (direction is forwards, but it's a bit off center).

Does it matter where MechJeb is on the craft, as long as it's pointed in the right direction? I have a large craft that tends to tip over at around 20-30000m during launch, even though center of gravity and center of thrust are perfectly aligned and I have plenty of RCS as well as gimbaled engines. I really want to do a successful launch with this beast.

Performance. It's annoying when one can't do a really epic mission because the game becomes a slideshow.

I've done a variant on that theme for my interstellar drive stage. I simply used two 6-way intersections and a small tank on top of them. On the 4 open connectors on each intersection, I attached an orange tank with a decoupler, with a few sepratrons (not sure if they are actually needed, but looks cooler). In other words, I had two stacked "plus" signs of orange tanks, eight tanks in all. Then, I placed, using girders, four nuclear engines (the slightly larger ones from NovaPunch) on the small top tank, so that the engines were in between the tanks. An X on top of the +, so to say. Then it was simply a matter of hooking up all the engines to the top tank, and then connect the orange tanks with fuel lines and arrange staging so that they were dropped two by two. Worked nicely, can take you just about anywhere. Another benefit is that, since it's flat and wide, less torque is needed to turn it, and RCS placed at the end of the orange tanks will get better torque from more leverage. Since it's also fairly short, it's easy to fit one hell of a lifter under it, and still house it in the VAB.

What are the main performance hurdles if one wants to put together a (cost effective, of course) computer for KSP? Raw CPU power? Graphics? Memory? Disk speed? I have two options at the moment: * A laptop, 64 bit with Win7, 1.6 GHz dual core, 8 GB memory. OK graphics, but not exceptional. * An older HP workstation, dual CPU with two dual core 32 bit Xeon CPUs running at 3 GHz. 4 GB memory, faster disk. WinXP. Not very impressive graphics. Now, I realize that neither dual CPU nor dual core will do much for me, but will the higher CPU clock and the faster OS make the older workstation outperform the laptop? Will the memory be a problem (I could add up to 12 GB, but then I'll have to run Linux on it, and Linux don't really get along with the multi-monitor setup on that machine.)? Which would be the best bet? And, for the general case, when one buys a machine, what should one aim for? Where's the important bottle necks?

Both Orion and Novapunch will lift a 200 ton payload into orbit, no problem. You'll probably need to use a bunch of Quantum struts to keep things from collapsing, though.

Check your fuel lines. Chances are that the tanks are drained unevenly.

OK, then I'll have no problem finding one. It's an airship with a deployable ground base, 8 tiny rovers, 4 tiny airship rovers, 1 larger airship skycrane/rover and 2 mobile labs (all at some 180 tons, iirc)...

Ah, then I'll have to add a command pod facing upwards. Is it important that it's centered over the the rocket, or does it just have to point in the right direction?

Kerbal Crew Manifest will help somewhat, by allowing you to transfer crew between modules in the ship. Then, once they are in the correct module, just exit.

When MechJeb is in control, what's "up" on a rocket? I have a rocket that tends to tip over, and I don't know if it's imbalanced drag or MechJeb thinking that up is towards the horizon. The rocket has an airship payload, so the command capsule is facing sideways (as is the MechJeb). Does it still know that up is the end that points away from the ground on the launchpad?