damerell

Crikey, I thought my old Mk IV was a monster. :-)

In the past people have had trouble with parts shifting on rovers during long expeditions. It may not be your imagination.

Another few false starts - however, one of them was caused by finding out I'd taken two engineers on the rover and left a scientist on the Hangarmoth. Having returned for a crew exchange, the ISRU situation is - still bad, but not as bad as it was. However, I think before I do any more roving I must plan a resupply mission to Duna with some additional drills to attach to the Hangarmoth. My other problem has been distinguishing smooth terrain from rough in the dark. Here I am experimenting with Scansat's slope mapping; however, there's no way I found to get it to show me much about intermediate states, just these blobs of red and green. The next idea was to plot altitude as discrete colours, not a gradient. Much better! The contour lines thus formed give an immediate idea of the slope. Here, for example, I'm about to turn to the north - I want no part of that cliff up ahead. And here I am having travelled a full 20 degrees longitude. Slow progress, but I think I'm getting a grip on how to go faster - and when the sun comes up (or rather, on Moho, when I drive under it) things should really ease off. Edited to put the descriptions over the pictures like the rest of the thread.

The Kerbal Foundries mod includes KSPWheel, an entirely different wheel simulation module, which was written out of frustration with the behaviour of stock wheels. (There are some KSPWheel configs for the stock wheels, although they're pretty old and may have rotted). Roving has many frustrations for me, mostly involving cliffs and the dark, but my wheels do more or less obey me.

I discovered recently the challenge rules have changed and my old scheme of using RCS for attitude control was forbidden. I decided to make a virtue of necessity and start afresh, redesigning the ships and rover a bit to reflect what I've already learned (one of which was "just send out a bunch of rovers to meet you en route"...) Nothing much changed with the Queen Agaster here except the inclusion of six Orion magazines - which made her much, much heavier but gives me a sensible reserve of delta-V once it's up. Bigger Ore and Metallic Ore drills (they're still not big enough) and huge internal storage tanks for them. The sticking point, however, is that some idiot (me) really nerfed the Convert-O-Tron to prevent over-unity operation (you can't dig up Ore, convert it to LF/O, burn that in a fuel cell, and get out more EC than you put in) and we're going to recover only tiny amounts of LF/O while landed. Undocked before landing on Moho. To add to this problem, a truly terrible docking maneuver (don't try and dock after a few kerbeers) means that we're almost dry on LF/O. I may have to send a resupply mission to Duna. Landed. In the dark, of course, and on Moho there's no waiting for dawn. And ready to go. The Hangarmoth is behind us with drills going to make what it can while landed (and, given the supply situation, I'll probably stay down until the next launch window or until habitation issues force a takeoff). But still, driving again, and even if the ambitious plan to circumnavigate all the things in one go fails, I might actually circumnavigate Moho. I've made some rover improvements too, but that'll have to wait for dawn.

Hm. My (currently in transit) Laytheboat uses Sink Em All (a mod) to let me pump ballast into the ship, then pump it all over to one side, which in testing seems sufficient to roll it back if it turns turtle.

Gosh. How does it transition from land to upside-down boat configuration? That's really clever for an amphibious approach.

I'm necroing this thread because I've pointed more than a few people to it over the years, because it's still the thread you hit if you try and search this forum to understand this option, and because reading this thread is very confusing until you realise that the left and right ends of the slider _changed ends_ between 2016 and now. What NathanKell wrote is entirely right - but if you just skim it, you'll move the slider the wrong way. What flart wrote is just quoted here to make it clear the slider has flipped ends; they make the common mistake about one setting or the other making the Kraken more likely. Here's how I understand it (and AFAICT this matches up with NathanKell): No matter what this setting is, every physics tick represents 20ms of in-game time. No change to this setting alters physics behaviour. (Physical time warp makes physics ticks represent more in-game time, and that can result in different physics behaviour). If the setting is low (in _recent_ KSP versions, the slider is to the left) and your computer is too slow, game time will slow down relative to real time (yellow clock) rather than the frame rate being reduced. If the setting is high (in recent KSP versions, the slider is to the right) and your computer is too slow, game time will stay the same as real time but the frame rate will reduce. Of course, there's limits to this: if the physics tick alone takes more than 20ms to simulate, the game time will slow down no matter how low the setting is and how fast your GPU is; if it takes 40ms to draw each graphics frame, you'll never get more than 25fps even if you have a one-part ship sitting still.

Moho's not completely devoid of scenery, but also, using altimetry scan data to avoid huge holes in the landscape doesn't work so well when the landscape is entirely made of huge holes. Another driving error knocked the front of the rollcage off (still, it died doing what it loved); too fast down a steep slope. Here I am at shutdown - another 18 degrees of longitude gone. Slow progress in difficult terrain. I think it's pretty clear the idea - even with KIS/KAS spares - of taking two rovers around all the planets was hopelessly optimistic. I'm not 1/4 of the way around Moho and two bits have been knocked off already, one of which I don't have a spare for. I must design a Mk VII (less prone to rolling, I say, as if I didn't say that every time I design a rover) and ship out spares.

An addendum: in practice I find it's going to be necessary to ship replacement rovers out in any event, since I keep knocking bits off them. Please can I get the one that's driving around Moho now grandfathered in - used to the old rules, didn't know they'd changed - and I'll make the replacements compliant with the new rules? ETA again - this request is moot, I decided I'd give up a few hours' driving on Moho and redesign the rover and transport vehicles anyway.

In this particular case it is a little awkward since I have just set off on an extensive journey with two rovers which have RCS and no reaction wheels (partly because I don't like using KSP's magic reaction wheels) and because I've discovered an ideal way to use MechJeb Rover Stability Control for landing attitude is to build such a rover and toggle RCS on when I want it to do anything. (The latter part of this is, I think, a pretty useful approach for other people.) AFAIK we're already dependent on people telling the truth about their journeys. Yes, if you say "RCS for attitude only", I could lean on RCS Translate Backwards every time I'm about to crash and not mention it, but if I'm willing to do that I could just turn on Unbreakable Parts and crash as hard as I want. (Also, if you'll forgive me saying so, the most original and amusing thing I've seen in Elcano recently is a journey around Gilly on a nuclear pogo stick - a journey composed _entirely of jumps_.)

I notice the rules have changed a bit since the 3 previous editions. What worries me is the addition of: Thrust devices (rockets, jets, props, RCS) are acceptable, as long as they are only used while in contact with the surface. Using thrust of any sort to control or change the trajectory of your rover while above the surface is not allowed. Now, I appreciate you don't want people boosting jump distance with rockets or RCS, but to use them to hit the ground wheels-down seems entirely normal to me. I certainly did it in my Mun circumnavigation (and mentioned it in the first post) which Fengist didn't even blink at, and since monoprop is finite it's arguably a more challenging approach than using KSP's magical reaction wheels. (Likewise, I don't have to look far to find Gilly trips where ion engines are used not just to stick the rover to the ground but to put it back on the ground when it leaves it.) The intent is that you control the vehicle during the entire circumnavigation. Using an autopilot goes against the spirit of this challenge. Fengist's original challenge says "You may use Jeb if you wish for it's 'rover' heading controls. If you read the account of my circumnavigation, you'll find an alternative to Jeb" and indeed when I did Kerbin for some of the long sea legs I programmed a heading into MechJeb, turned on the engines, and went to bed. ETA: now, I ask myself, why do I think that was OK but some rover autopiloting wouldn't be? I think largely because of my own no-quickload approach; if I got the heading wrong, underestimated the speed of sailing, or just found out MJ wasn't a very good steersbal, it was up to me to deal with the consequences. (I'm less personally concerned with this change since now Scatterer has real waves, there's no prospect of MJ handling sea legs.)

"Ship Manifest" has a bunch of cheat tools to create or modify Kerbals.

Moho hasn't really had much in the way of wow scenery moments, so here's just a shot of the rover as I shut down for the night IRL. I am covering about 20 degrees of longitude an hour, which... well, I knew it would be a long trip when I set out. I've also - oh, no, not another learning experience - found that letting the rover run away downhill is dangerous; kOS brakes it at 43 m/s (otherwise the wheels will fail) but it's very easy for it to skid then, Moho's light gravity produces a huge spin, and when it gets side-on it can roll. I knocked off one of the solar panels, but they are outside the roll cage. I hope this is just teething problems, rather than indicative of the rate at which I can expect to make driving errors.

More comedy; clearly whoever planned the transfer for the Moho scanner probe was more sensitive about dV usage, because it's arriving 15 days after the kerballed mission. Ooops. Fortunately, I put a set of SCANsat parts on the Queen; she's parked in an equatorial orbit which will scan the area I actually propose to drive around. Here I am landed on Moho: I had tested the ship lander script in development, but it's still a relief to be down and dry. The Hangarmoth guzzles LFO, MP, and pulse units alike - I'll see to what degree ISRU can help with that. (Bear in mind CAMREC very severely nerfs Ore conversions to prevent over-unity operation, and it just takes a lot of drilling to make pulse units). Here I am about to set off: And (after a few false starts) here I am in daylight, 27 degrees east of where I started. Moho's long day means the sun didn't so much rise as that I drove under it. There's no avoiding driving in the dark on Moho - the life support and habitation timers won't let us camp out for weeks at night - but keeping it down to 20 m/s or less let me pick my way through without incident. Now the sun's come up I can put my foot down a bit. I've also made two improvements; now the Queen has passed overhead and scanned the terrain, a zoomed-in altimetry map lets me get an idea of where there are craters not to drive into, and since the rover has no enabled reaction wheels (I'm not sure it has any at all) I can leave MechJeb Stability Control on and it does nothing until I press the RCS button. This is a much better idea than trying to detect spins with kOS - the button's literally under my right forefinger so as soon as I think there's any danger I can have MechJeb level me out. What I mean by "false starts" is, the way I do these Elcano missions is no quickloading during the actual drive. I can quicksave and quickload getting to the start, but if I damage the rover I can either press on or start again from the beginning. I took a couple of goes to get the hang of driving the rover in the dark, and that's allowed, but I can only load the save at 147 degrees East once. (Obviously, this doesn't apply if KSP crashes, etc; but driving errors are permanent.)

I've never used timewarp during circumnavigation other than as necessary to get up to 100% of normal time. That said, on Kerbin I did program in an ocean course and go to sleep, but the rover sailed itself in real time...

It strikes me at some point we should have a picture of the rover. The Mk IV that circumnavigated Kerbin was not really built with any regard for its dimensions at all - after all, it didn't have to fit in anything else. It also predates CAMREC - ie, it was a 55 tonne electric vehicle you _could_ power by slapping on a few solar panels. The solars on top of the Mk VI are just there to keep life support going if there's otherwise a total electrical failure. I briefly experimented with a half-track Mk V - keeping the Karibou cab and one set of mole tracks from the Mk IV - but then thought, well, why not go the whole hog and dispense with the huge tracks altogether? So, the Mk VI is a six-wheeler: USI Life Support also complicates matters by making 3 kerbals not want to sit in one cab for days, so there's a second crew part just behind it. A side airlock and ladder provides a second exit point - there's spare ladders in the KIS/KAS spares locker under the vehicle, but if you only have one ladder and you knock it off, it's still really inconvenient. Most of the rest of the rover is devoted to electricity; a nuclear reactor (which can't manage the peak power output but can run forever), large batteries so the reactor can top them up when the rover isn't being driven hard, two large fuel cells to supplement the reactor when battery levels are low, and a very large lump of LFO storage for them. There's also two small engines to assist on very low-gravity worlds, and RCS to be able to orient the rover on large jumps. This seems a good time to write about kOS support for the mission. There's a kOS script that runs those fuel cells and brakes on overspeed, but also detects a high spin rate or being at too high an altitude and tries to level us out with RCS before we hit the ground again. The Hangarmoth has three scripts; one to balance weight (simplified by the assumption we can probably do everything we want by moving heavy atomic pulse units around, and improved over most fuel balancers by the realisation I only really care where the centre of mass is in the plane perpendicular to the thrust axis), one to assist maneuvers (by dialling down the pulse size to make the last pulse "just right" then activating the auxiliary engines), and a script to land it on airless worlds. Here's the Hangarmoth + Queen Agaster assembled and setting off for Moho: Just after taking this I realised a) the Duna scanner was probably going to arrive on Duna after I did and b) the Eve rover descent and Eve ascent vehicles... hadn't been launched yet. Thankfully, there was enough time to set them up with some slightly dV-expensive transfers and with any luck that'll all be fine.

Three ideas simplied this. One was to abandon the spaceplane rover lander idea altogether. There's only two other worlds with significant atmospheres, and Laythe doesn't even want a normal rover landed on it (you can circumnavigate it entirely on water, but not on land). Special arrangements can be made to land a rover on Eve and a boat on Laythe (which just get abandoned) and to return from them. Now, the rover lander still wants an end-on configuration to take off from Kerbin, but we also now have no reason not to take off vertically from the launch pad. It still wants to transform into a belly landing configuration, but that's much easier to do if we don't have wings on the side of the ship just where we'd probably want the Orion units to attach. The second was DockRotate (which predated the stock docking port rotation, and works a lot better). Instead of guzzling RCS fuel to reconfigure for belly landing, the drive units can easil y be rotated into position. The third was to have a bit of fun-with-Blender to use the (otherwise ideal) LLL drop-ramp cargo bay as a Hangar; instead of the rover adding a huge number of parts and having to be secured to the lander manually, it just becomes a lump of mass as soon as the cargo bay closes on it. Also, the bay TweakScales in awkward increments and was rather too large - but as a Hangar, we can make a virtue out of that and bring a complete spare rover (and a couple of very basic seats-and-engine vacuum taxis for rescuing kerbals from unwise EVAs). This gave me the _Hangarmoth_: I'm going to go easy on launchpad shots of ships which are just boosters, engine, tank, fairing; but the Hangarmoth is a bit more interesting. The huge boosters handle the initial launch so we don't have to rebuild the radioactive ruin the Orion units would leave behind (and as a bonus this saves a few atomic pulse units). The Orion units take over about 10km up and provide almost all the rest of our propulsion; there's an auxiliary LFO engine outboard of each Orion unit for finetuning of maneuvers (and also the case where we're approaching something we want to dock with too quickly and want both to slow down and not to blow it up). The fins provide a bit of control on the ascent but are discarded when we get to vacuum. Besides the cockpit and KIS/KAS storage, the forward portion includes an MKS Tundra centrifuge to produce enriched uranium and an MKS workshop to produce metal from metallic ore. (The Orion spinal parts can produce additional pulse units from these, and there's one inside the ship). We've drills for both of these and Ore, and a Convert-o-Tron. I don't know how well this is going to work out in practice, but it's my hope the ship can resupply itself with pulse units, LFO, and monoprop; it is after all going to spend plenty of time sitting around on planetary surfaces. The main body of the ship has the Hangar in the centre, two large cargo bays (mostly containing pulse unit storage but also a nuclear reactor), and two very large landing legs. Here it is having gone to space today: And here it is DockRotated into the belly landing configuration: Its first objective is to rendezvous with the _Queen Agaster VI_, the life support module for the mission. She was lifted into orbit by a normal boosters-etc approach, and is seen here about to circularise and deploy her centrifuges: And here she is with everything deployed and ready for the rendezvous: The Stockalike Station Parts centrifuges provide enough habitation time for the mission. Three MKS Tundra recyclers at the rear will make our Supplies go further (the fourth Tundra module is a medbay which I don't expect to need but it makes things look tidy), but most of the life support supply is two large Kontainers of Fertiliser which agronomics modules will use to convert Mulch back into Supplies. We should have enough endurance for the entire mission, but if we don't, we can always send a conventional mission with resupply to one of the later worlds. There are a total of 12 kerbals on the mission; three pilots, three scientists, and six engineers. The expected typical landing is pilot/scientist/engineer/engineer; one engineer will stay to run the drills on the Hangarmoth while the other three drive around something. This leaves eight in orbit on the Queen Agaster; RemoteTech wants six to provide local control of probes (ie, without the time lag to KSC).

I'm not sarbian, obviously, but I'm really not sure what you're asking. If you want to start a thread, no-one is stopping you. You don't need permission from anyone if it's in line with the forum rules. If people use it, great. If not, well, it turns out it was easier to keep stuff in this thread.

It may have been a little while, but I still want to circumnavigate all the things, and I've reached the point where I've designed all the vehicles I think I need to launch from Kerbin - indeed, the ScanSat satellites are already in transit. I don't want to use the absurd design above; it only works at all because the normal inflatable balloon in the Hooligan airships mod produces unrealistic lift. There's one version of it in the current plan, but it's the realistic lift version - it can lift a bit of rocket off Eve at the cost of making a horribly unwieldy object to get onto Eve, not two kilotonnes off Kerbin. This time around I've the following mods to make life hard: USI Life Support (you can ship out Supplies to Kerbals, but you can't stop them missing home if you stow them in tiny lander cans), RemoteTech, FAR, and my own Campaign for Real ElectricCharge which means there's no driving a 50 tonne EV around a planet by sticking three solar panels on top. Conversely, to make life easy I still have Project Orion in the form of SuicidalInsanity's mod. The Orion spaceplane design returned in the shape of the _Behemoth III_: This was no more practical than the Behemoth, but the pieces were coming together; the LLL drop cargo bay from the Behemoth, the twin Orions, the idea briefly considered of undocking the engine modules and RCSing them around to dock to a ship much more suited to landing on worlds without atmosphere.

Your adventure on Dres reminds me of getting home on Kerbin, where in the very last bit of the journey the steering suddenly and inexplicably got stuck and I had to limp over the line with MechJeb and me fighting the controls. Only, presumably, for rather longer.

Suppose they do; suppose literally everyone but you and me stops playing KSP1 immediately. This isn't the end of the world, because every mod that works now for KSP1 continues to work indefinitely. (Linuxgurugamer is also maintaining an astronomical number of mods and has a Patreon...)

Kind of weird getting that on Minmus - ah, stock wheels. With KF wheels I'd expect them to recognise there's almost no traction, so while the vessel being light would make it easy to nose-up, the wheels just wouldn't be able to apply very much torque. Alternatively I'd have switched to FWD so that the instant the front wheels left the ground there'd be no more drive torque.

Likewise. The stock headlights are near-useless. However if you are using a life support mod it may not be practical to only drive during the day, especially with Moho's very long solar day. I found the MTS light from https://spacedock.info/mod/2369/TrackingLights makes a reasonably good "big headlight" (indeed it can be turned up to provide truly absurd amounts of light); I've got two on my Mk VI rover. It's considerably smaller than the Spotlight shown on the mod's spacedock page.

Silly question, but for Kerbin couldn't you just drive it out of the SPH?