Tiron

I used that craft in a video I posted earlier to demonstrate the core issue. Compare this pre-release footage: to this 0.21.0 footage: I'm doing the same thing he did, giving it some rotation (just on yaw rather than roll because it's easier to quantify), and then letting the SAS have it. It's still doing the same thing in 0.21.1, just now it stops sooner (65 degrees rather than 90 when holding full yaw through 45 degrees of rotation), and rotates back towards the release point more (10 degrees as opposed to a negligible amount), but it's still basically doing the same thing. Edit: I'm going to a do a 0.21.1 version, comparing mechjeb's killrot to the default SAS, and clearly demonstrating that it is NOT a Deadzone issue...in just a moment when I finish primary mapping of Kerbin.

It's less of a 'fix' and more of a 'mitigation'. The actual problem got mis-attributed to a joystick deadzone issue, and went untouched so far as I can tell.

This is what should be happening: You'll note that when he rotates the station and releases it, it then stops the rotation, then brings it back to where he released it at, and damps it down there. Exactly like Mechjeb's Killrot would do. Everything I've seen and heard indicates this is the intended behavior. What's it's ACTUALLY doing is not locking onto a heading until quite a bit after you let go, and then holding that. It causes 2 things: It uses less force to damp the rotation than it's actually programmed to, because the incorrect 'target' is closer to its current position than it would be if the targeting were correct. The second is that it keeps sliding for a ways after you let go before locking onto a heading, making it much harder to get it on the precise heading you want. The 0.21.1 changes make it lock down MUCH faster, so the deviation is lessened and more noticable, making it easier to correct. And once you do get it aligned, the extra force it applies makes it hold that attitude much more strongly than it did in 0.21.0. It's better, it's still a bit tricky to use. I'll be making another vid to illustrate what I'm talking about, with an extra bit to make sure there can be no doubt at all it's not a deadzone issue.

The Core issue of the target heading not being set properly DOES still exist. It's just now masked a bit because the 0.21.1 changes make it much easier to compensate for. It is a LOT better than the 0.21.0 implementation, but basically harvester is so convinced that it's a joystick deadzone issue (even though there's large numbers of people who don't even HAVE joysticks that were having the problem), that they 'fixed' that instead of actually finding the real problem. Yes, it's much better. Yes, it's much easier to use. As it stands, it still has precision issues that require a number of adjustments to compensate for. It's just fewer, weaker, and less frequently. If they fix the attitude hold targeting it'll be better than Mechjeb's Killrot. Right now it's a toss-up: The default SAS is much more efficient and less wobbly, but Killrot holds a heading better and requires less manual input to get it and keep it on a heading in the first place.

If they had low 'stupidity', they won't know what to be afraid of and what not to be, so they'll panic at things they shouldn't, even with higher brave Stats. Example: Bill and Bob. Bill has 'Stupidity' 0.8, bob is 0.1. Their 'Courage' ratings are 0.3 for Bill and 0.5 for Bob. Despite being braver, Bob screams more and almost always looks worried, because he's so bloody dumb that he doesn't know what's going on. Bill has a pretty good idea what's going on and so doesn't scream nearly as much, but he's so cowardly that ANY potential danger at all sets him off. For less screaming, you'd want high on both.

Three Things: Mechjeb. Smart A.S.S. KillRot.

I'd still like to see the attitude lock not setting properly get fixed, as it'd take it from 'very good' to 'absolutely glorious'. The new settings mask it enough however that it's fairly usable in its current state so a fix for that can be a lot less urgent. it's still not QUITE right though.

It's still got the attitude lock problem (apparently they didn't see the video I posted before or something), but it's a LOT easier to deal with now. I mean, dramatically easier.

In his case it doesn't actually matter what he's set to: His BadS flag overrides his stats and makes him grinny almost all the time.

It's better, the much faster and snappier response helps a lot in dealing with the attitude lock not setting properly. It's still there, but it's easier and faster to deal with, and I think a lot less prone to losing control of crafts.

Well my joystick wasn't even plugged in all this time, and there's a lot of people having it that don't even have joysticks, so unless there's phantom inputs, it's not that. What I see is that it's not setting the attitude lock target at the point where you release the controls, but setting it at some point after that, the distance it's off being proportionate to the rotational speed. My speculation is that somewhere in there the SAS itself is retriggering the 'Change Target Attitude to Current' function. It's definetly setting an attitude and locking onto it, eventually, it just seems to do so, for the final time at least, really late for some reason.

I used to just design a lot of safety systems into my designs (no decouplers for the cockpits in the B9 Aerospace pack is one of the main reasons I don't use it), and do some cheatyness if something goes really wrong. Now though...Unmanned test flights of a manned design...or just use the cannon fodder instead of the prime crews. (IE: I Hired a bunch of dumb ones to use on really dangerous stuff).

I'm guessing it wasn't so much that only some people were having it, as only some people were BOTHERED by it. Anyway, new testing results: It's MUCH snappier and far more aggressive about slowing the ship. In my spin test with the Kerbal-X final stage, with a 45 degree initial input, it's now only taking 65 degrees to get it hauled in...and then definitively snaps it back a bit...but still only about 10 degrees. It's coming to a stop 55 degrees from the point where I let go of the controls, so for gross movements it still needs quite a bit of help. Trick being fine movements are much improved. It's still a bit tricky to get the attitude lock set on the precise point where you want it, but because it's so much snappier and aggressive about holding it, it takes far less time and many fewer adjustments, and it's actually quite good at staying there once you get it there. It's also much more immediately noticeable when it's got the lock target set a little bit off and pulls it off, because it's a much more rapid movement now and happens sooner, so it's a lot more obvious when you have got it there (or not). Not quite perfect, but I think much more usable now.

Oh aye, I was just saying. On the SAS Front: Pretty much exactly what I thought was going on from my test. There's also been some other tweaks to it too. I'll be taking a Kerbal-X up in a second to test it, but it sounds extremely promising.

Let me add in Atmospheric entry simulation. Mechjeb is literally the only mod I know of that's able to predicting a landing spot accurately through an atmosphere, and the same function can accurately estimate your orbit after aerobraking. They're both extremely useful and I know of nothing else that provides it. But the basics is this: All of Mechjeb's autopilot functions have to be turned on by pushing a button that turns it on. Each of them also has all sorts of informational cues and prediction that goes with it, that you can use while flying manually. Plus all kinds of other, detailed information you can't get out of the default UI. Using the navball's target to indicate a given ascent path, or the glideslope for landing on the runway, for example. I firmly maintain that the spaceplane autoland feature is terrible (It once bounced a plane, then turned it around backwards to point back towards the landing point, resulting in it landing backwards and shortly being rendered into pieces tumbling down the runway) and refuse to use it, but I use the ILS guidance every time I land a plane (and sometimes ignore it if it's being stupid). Even if you don't want the autopilots, it's worth using just for the informational and predictive stuff.

Bill has High Dumb/Stupidity (0.8), Bob has Low Dumb/Stupidity (0.1). Which one is the complete moron, again?

Unless it's been changed... it's BadS And yes, only Jeb has it so far, although prior stuff from the devs has mentioned 'pilot training' being one of the factors that affects how the Kerbal behaves (and implied it was referring to the BadS flag). And interestingly, it's been this way in the stats forever behind the scenes but now it's showing up in game as well: The 'Stupidity' stat is backwards. It's more like an 'Intelligence' stat, because a higher value indicates they're SMARTER. What that one does is illustrated very profoundly by Bill and Bob. Neither is very brave (Bob maybe a tad more than Bill), but Bill has high 'Stupidity' whereas Bob's is very Low. What this means is that Bill has a pretty good idea of when things are actually dangerous, and usually only freaks out accordingly. Bob on the other hand doesn't have the faintest idea what's going on and is scared/worried basically all the time, unless it's flat out obvious it's safe (IE: Floating under a parachute, on the ground).

Well C7 has said that they're looking into the cause of the problems a few pages back, so I'm pretty sure the added-difficulty isn't intentional. and the current behavior isn't what was seen in previews either, they were showing an attitude hold closer to Mechjeb's Killrot than the old ASAS, only with vastly improved use of thrust and much more stable.

Last I heard this was a steam thing. If anyone's got this on a squad store version they should probably speak up about it. As for me, my fresh, clean steam install actually ALSO has them in gamedata. So I'm guessing it's just a steam frak up. I'm gonna try deleting the ones in the old parts folder and see if it still works, I'm guessing it will. Edit: Seems to work fine if you just delete them. Seems like Steam just being dumb again.

Well Originally, the Command Pod's SAS system was set up so it could only damp motion and not actually hold a heading, rather like what we're seeing, because it was supposed to be a demo version the Kerbals got trying to sell them the full version. Problem is, they're now ALL doing that, in a weird way that I think is caused by it actually being a full attitude hold, except the 'target attitude' keeps resetting. I've heard secondhand(at least) that the devs said it's no longer PID based at all. That said, the way it's acting on mine is very similar to how the old PD-only ones were working. Only slower, and with a brief control reversal at the end of it settling in (as if it were starting to go back the other way to reach its target point...which is right there, so it only does it for a fraction of a second)

Well see, there's something of a misunderstanding going on. People are shouting for the old SAS back because it could actually *hold a heading*, not because it works better. Anyone who thinks it works better is daft. What I'm sure almost all of us actually want is the new SAS working like it did in C7's videos, where it basically works as a better-in-almost-every-single-possible-way version of Mechjeb's Killrot Function (which is itself a vastly improved version of the old ASAS). Right now it seems to be working as a better-in-almost-every-possible-way version of the old Avionics Nosecone.

It's too widespread, and from what I've seen of it, far too basic to be isolated. I think some people are just not bothered by it as much as others. Basically, it's not acting as an attitude hold at all, it's acting as a motion damper, and one that's very slow to kick in at that. It makes it almost impossible to hold a precise heading in space, because there's no drag helping it out. On a spaceplane it works spectacularly, because it's basically working as a drastically improved version of the avionics nosecone. But it's doing that in space as well, where you really need it to act like Mechjeb's Killrot.

Actually, it's not so much 'hating' as 'it doesn't do the one thing we most need it to to do'. It's SPECTACULAR in every way except that it's merely a motion damper (and a slightly slow one), instead of an attitude hold. Based on pre-release material, it's perfectly capable of functioning as an attitude hold in a manner similar to Mechjeb's Killrot...only much, much, much better in every possible way. It's just not doing that at the moment, making trying to reach and maintain a precise heading difficult in the extreme. I'm praying desperately it's a bug, but I just woke up.

Ummm...it'd been in for quite awhile, it just wasn't so obvious. The names, stats, and statuses of all crew members has been stored in persistent.sfs all along, there just wasn't an ingame interface to see it. If a kerbal was killed, their status code was changed to '3', only used for dead kerbals, and their 'ToD' field was changed to the time that they were killed at. They still stayed in the persistence file, though, even as replacements were added. The functionality itself really hasn't changed, just the particular implementation of it, and they've added a UI that lets you see the stuff that was already there. I'd been looking at it in notepad anyway... I do manual persistence and craft file hacking occasionally, usually to fix something that broke. (IE: Earlier I hacked a craft file to remove mod parts from it, to determine which one was breaking the game. It was mechjeb, as it turned out, since fixed). Edit: Just to point out, the only really 'new' thing that isn't just a UI, mechanics-wise, is that it gives you a list of kerbals and lets you pick which ones to put on the flight team, instead of just automatically generating a new one when it was needed.

In a way it's not. Those videos showed a system that worked like Mechjeb's Killrot has for as long as I can remember: Controls override it, and when you let go of the controls, it sets that to the new attitude hold point and holds it there until you move again (or turn it off). What we got is a system that works like the old Avionics Nosecone or Pod SAS: It merely damps movements, and doesn't try to hold a particular attitude at all.