TimTheTerrible

April fool!!!

@AlexinTokyo: Thanks, that's it!

@Amaroq: I didn't know that about the SPH, thanks! I haven't spent a lot of time there. @Red Iron Crown: Yes, reorienting the original part (the octo) fixes the problem, but unfortunately, that makes the rover challenging to attach to a landing vehicle (attachment balls oriented the wrong way). But leaving that aside, I found something interesting: I have the Bill9000/Jeb9000 mod to go with mechjeb, and where I stick the Bill/Jeb module and how it's oriented changes the orientation of the vehicle as well: if I place a Jeb9000 on top of my original "broken" rover as shown below, it fixes both the compass heading problem, as well as the autopilot weirdness. Interestingly, the original mechjeb AR202 module doesn't exhibit this behavior: no matter where you stick it, the orientation remains "broken." (FYI: additional parts in image are from FASA Mercury/Gemini kit) Finally, now that you kind folks have helped me understand how to make working rovers, can somebody (Amaroq?) tell me how to actually change the title of this thread to [Answered]? Is it as simple as typing "[Answered]" in the title of the advanced post page? I can't seem to find a control to do it. Thanks all!

@Red Iron Crown: Oh, but of course, it didn't. ;-) Poor quality observations on my part, for which I apologize. Now that I've had some time to tinker with it, I can see how I thought it was working: the ground around KSP is more or less completely flat, so operating in that area, under just the right circumstances, it seems like it works. More rigorous testing showed me that it didn't actually work, until I rotated the control module 90 degrees forward.

@Kasuha: I stared at that picture for a whole 2 microseconds before the light came on over my head like a supernova. It's perfectly obvious now why it was doing what I saw it doing. That does, however, raise the question of why the stock rover, provided by Squad, looks like this:

I'm using a probeCoreOcto as the base of the rover, it's right-side-up and the nav ball shows all blue, and I'm driving in docking mode (linear) with the reaction wheels shut off. I have no trouble at all controlling the rover, it's just that following a particular compass heading (0 degrees = North; 90 degrees = East; etc.) doesn't seem to work: if I aim the rover at 90 degrees, for instance, and drive over perfectly flat terrain for a minute, the heading will wander off to some other reading, even though I'm still headed in the same direction with regard to the local visual reference frame (which is to say, pick a conspicuous land mark and head toward it). I'm well aware of the fact that, starting from the North pole, every direction you go is South. The effect I'm describing is seen even over extremely short distances (under 1km). That's why I asked about magnetic anomalies.

@Red Iron Crown: When at rest, the navball is all blue, with the heading mark in the center, so "right side up." @cantab: I got bit by that one at first, couldn't figure out why the rover kept flipping, or as Bilbo Baggins would say, "like riding a fat pony who always wants to roll in the grass." So, yes, I switch to docking mode and disable the reaction wheels. It's not that it deosn't go where I point it, it's that the thing that's supposed to tell me where I'm going doesn't seem to give the same advice each time I look at it. I'm at work now, so I can't easily test (even though I work at a video game company!) but when I get home tonight I'll do some formal testing and see if I can characterize it better. Thanks for the replies!

So, I spent a lot of time lately trying to do long-range rover science on Mun. I started out trying to use the MechJeb2 rover autopilot but eventually gave up in disgust and went back to driving the rover myself. After quite a lot of "Now where the fark am I?!?" I came to the conclusion that the navball heading indicator doesn't seem to work reliably on Mun. Incidentally, I haven't tried to navigate on any other body using the navball heading indicator, so I don't know if it does or does not suffer from the same apparent problem elsewhere. While motoring around Kerbin the navball heading indicator always appears to be correct and reliable: head toward an object, get a heading fix; wander around in circles for a while, then go back to your marked heading, and lo and behold! you're once again heading toward exactly the same object. Try that same trick on Mun and you'll end up doing a Dukes of Hazzard off a crater rim and lighting up the dark side with a fireball worthy of JJ Abrams. Fun to watch, but not a lot of science gets done. Is this some intentional aspect of game design (magnetic anomalies, maybe?) I'm not understanding, or is there a bug, or am I just daft? Edit: I have edumacated and can now build working rovers!!! Thanks everyone!