Come back old ASAS - all is forgiven!

[spikeyhat09]

its better. its not perfect (and i can't say that i prefer it to the old ASAS system), but its better.

[check]

Still not 100% satisfied with the SAS though. The probe's integrated SAS is too weak and the reaction wheels are too strong and heavy. Damnit.

It's just gonna need probe size reaction wheels, as well as large rocket sized reaction wheels.

The 0.21.1 update fixed most of my SAS concerns. Things fly much, much better now

[Tiron]

It's just gonna need probe size reaction wheels, as well as large rocket sized reaction wheels.

The 0.21.1 update fixed most of my SAS concerns. Things fly much, much better now

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.

[bogpad]

I just tried a basic 2.5 stages rocket of small/mediumish size, with 49 parts that was a bit tricky to get into orbit in .21.0, and now it works absolutely FANTASTIC.

not sure about really huge rockets with lots of parts, but for now, I say the new SAS works like a charm!

[Nicok]

I do have a theory on why some people were affected while others weren't. The drifting issue was related to how the SAS would release attitude lock when input was detected, and that could have been getting interference from joystick axes with zero deadzone.

We corrected that by setting a minimum deadzone that's independent of the joystick deadzone, so if you have joysticks with no deadzone, it will still properly ignore the jitter from the analog inputs.

I think those who didn't have this issue either didn't have any noise on their input axes (as detected by unity), or had deadzones set for their joystick assignments. Not sure... That's just my theory of how it could affect some and not everyone.

Cheers

The fix improves the IAS, but they are still some problems à think.

Firstly, without any joystick, i had the issue of 0.21.

Secondly, there is some kind of...waiting when you mark an attitude to go and the effective attitude hooked. It's like during a moderate rotation, you're not sure about the real attitude hooked. Can't we have some mark on the navball like "here is the attitude I'll looking for" message from IAS ? A way to know exactly where the IAS drives us...

[Pixel of Life]

Well, they said they are continuing work on the SAS so there's probably going to be another hotfix in the (hopefully near) future. But for now, this will do just fine

[DMagic]

I just tried a basic 2.5 stages rocket of small/mediumish size, with 49 parts that was a bit tricky to get into orbit in .21.0, and now it works absolutely FANTASTIC.

not sure about really huge rockets with lots of parts, but for now, I say the new SAS works like a charm!

Did someone say big rockets with lots of parts?

This works fine, it holds steady going straight up, it can make the gravity turn just fine, and it holds its course. It has just two inline reaction wheels, a command pod, one gimbaling engine and a bunch of winglets.

[Giggleplex777]

TBH, I actually just want the old ASAS back.

Edited July 26, 2013 by Giggleplex777

[SRV Ron]

Rebuilt the Tri Orbiter for 0.21. The SAS is working just fine now that I am learning how to control the orbital turn. Got the design into a good 90* plane orbit, then on to orbit around Mun where I will later land the probe there.

[Basilo1146]

Ok, my input. I'm sure this has already been mentioned, but i'm not reading through another 45 pages to check. What I have noticed with the new ASAS is that, while it will stop a ship from turning/rotating after manual input, it seems not to cancel out the effects of other forces. For example, if I have a spaceplane and I hit W for a second, it will cancel out the movement created from that input. However, because the tail fin puts the CoM ever so slightly out of line with the CoT, the ship experiences spin under acceleration. Rotation from sources like this, such as wobbling, drag, lift(or lack thereof), or unaligned CoMs and CoTs will not be stopped, which is a major problem. Another example of this would be that any deviations of course that occur when rolling/banking in the atmosphere, such as the nose dipping towards the ground, will not be stopped as they are not directly caused by manual input. Of course, I could be completely wrong, but this seems to be case in my game.

[dlmarti]

TBH, I actually just want the old ASAS back.

Except for docking, I would agree. Keeping on a maneuver node is a little tedious.

Docking on the other hand is pretty good, but then again I made a big effort while waiting for the release to get really good at docking... So maybe thats just me.

Unfortunately for the aero guys, this is a big plus, so I doubt we can get the old system back. They had it pretty bad under the old ASAS system.

I fear that since this is not a PID feedback loop, that it will never be perfect, so lets just hope for a little bit more improvement in zero g.

PID controlled feedback loops scare people for some reason, and almost every new engineer I get, wants to do it differently. It never works out. Granted they can be difficult to tune sometimes, but really what else is there? Most people do a simple proportional loop, and then kludge in a bunch of code around it. Then they spend 6 months trying to get it too work, it never does.

I learned on motor controllers, so I had the mass of the rotor to dampen my early mistakes. So I guess I had it easy.

[Benie]

TBH, I actually just want the old ASAS back.

I seriously need to try this update. I'm curious how rovers will behave with reaction wheels; like my Speed Rover. Also kind of scared to see the results as well.

But in one way I'm glad I was wrong for once. Looks like not all developers are all "don't like it? Tough!". I'm glad SQUAD isn't one of those. I now have respect for them.

[Tiron]

TBH, I actually just want the old ASAS back.

Three Things:

Mechjeb. Smart A.S.S. KillRot.

[Fernbhoy]

i declare this latest patch........

A SUCCESS!

Edit, enjoy that pic, i then proceeded to undock my ship with no rcs installed and ram off a solar panel trying to head back to kerbin

Edited July 26, 2013 by Fernbhoy

[Guest]

TBH, I actually just want the old ASAS back.

After trying the update, I'm going to have to agree. The 21.1 SAS just seems to have become a tighter version of what we had in 21.0. It still doesn't stick to the heading I set during a burn, it just drifts less from it. Once the burn is done is returns to the proper heading. It is definitely an improvement but the slightest imbalance in a craft completely throws off its ability to stick to a maneuver node.

So, question for the devs: Is the intention of the new code to trim the movement to a stop instead of attempting to maintain a heading?

[Giggleplex777]

I noticed something while I was trying to bring back the old ASAS. The new SAS overrides the ASAS' input and renders it useless. I had to remove the SAS from the probe core's config for the ASAS to work.

[_Aramchek_]

Three Things:

Mechjeb. Smart A.S.S. KillRot.

Or just fix the ASAS if indeed it is broken, not everyone wants to use mechjeb, or other mods.

Judging from .21.1 Though there was a problem with ASAS, and it is MUCH IMPROVED in the hotfix build. Thanks Squad.

[Benie]

To change something that everyone's been used to for months, just because it's more realistic, feels wrong. We weren't even warned about the fine print behind the "cheat mode" ASAS. That's why everyone has been upset.

Why couldn't they had made it an option for the future Career Mode, that you have to use this newer ASAS? That would make much more sense.

Just my 2c.

Edited July 26, 2013 by Benie

[Guest Aaack]

I just tried the new ASAS version, it saved the day.

Yep it's different to the old one. I'll need to adapt but it's really effective and good. Perhaps the issue is that is so transparent some people don't even realize the black magic is doing in the background...

(for details of the ASAS saving the day I wrote a mini report in the report section of the forum)

[boxman]

@HarvesteR

I dont even have a joystick plugged into this pc, but yet i oddly enough noticed a big improvement with a ship that was completely unflyable with .21.0

Either way thanks alot for fixing it this quickly.

[sanity]

v0.21.1 is a lot better, it actually works and i feel i could get used to it, but i would like an indicator on the nav ball so i could tell where the SAS has locked on too, cause sometimes its hard to tell

[Guest]

Think you misunderstand what I mean. I am not saying you do not have a right to complain, by all means go ahead. However, you might have to come to the conclusion that it may not ever be exactly what you want.

Then we need to know what the intended behavior of the new SAS system is supposed to be. If it's not what we want we can find other solutions.

[Brabbit1987]

Well hopefully they fix whatever is broken then, though I am unsure how the last patch fixed it for most and only a portion of you still seem to have the issue.

[Guest]

Well hopefully they fix whatever is broken then.

I'm sure they will, if anything is still broken. I'm about to do a fresh wiped-directory install from Steam to see if the problem is persisting. If it is, I will file a bug report.

[Danger Will Kerbinson!]

LOL I am not even going to comment if you want to misunderstand what I mean and act like a dick.

Oh, Im sorry...wanting an issue looked into is a dick move now. But screaming AGAINST a bug being fixed is some noble deed? What was wrong with you people that did NOT want the minority of us to play? I'm waiting in anticipation of this stunning revelation.