kOS Scriptable Autopilot System 0.9

[KevinLaity]

Except that I just discovered that support for this:

lock steering to up * V(1,2,3).

just went away. I can't use the V() expression with "up" or "north" or "steering" anymore in 0.6.

So now I'm trying to figure out how I'm supposed to convert from xyz distances to the new angle-based steering when the language no longer supports that expression, and it doesn't have the arc trig functions (arcsin, arccos, arctan) to convert from vector lengths to angles.

I'm getting the impression that despite giving us more information the removal of that expression "up * V(1,2,2)" may make it actually harder in the new version.

I'm still messing with it trying to figure out how to do what I used to be able to do.

In the code I've split vectors and rotations back into separate types as part of exposing :X :Y and :Z.

I had added an implicit cast that I thought would have automatically converted vectors back into rotations when used in expressions. It looks like I just need to go back over it and make it more explicit.

[KevinLaity]

I'm having issues with it pitching over for the gravity turn on my test rocket.

I'll take another look at the steering.

The reason it changed is because when a vessel is in orbit, it was wildly oversteering causing wobbles that last forever. It also did rotations extremely slowly.

I may not have tested it with enough variety of rocket designs. If I can't get it fixed I'll revert back to the old method.

Pitching over during a gravity turn is usually a sign of not enough torque but it could also be understeering on the part of the mod. Would you be will to send me a private message with your craft file attached?

[Payload]

I'll take another look at the steering.

The reason it changed is because when a vessel is in orbit, it was wildly oversteering causing wobbles that last forever. It also did rotations extremely slowly.

I may not have tested it with enough variety of rocket designs. If I can't get it fixed I'll revert back to the old method.

Pitching over during a gravity turn is usually a sign of not enough torque but it could also be understeering on the part of the mod. Would you be will to send me a private message with your craft file attached?

It is definitely understeering now. It wasn't over steering before. It was experiencing Reset Windup and Derivative Kick. Reducing control force globally wont fix those problems globally. Only locally.

Edited September 16, 2013 by Payload

[Gaius]

Since the amount of steering needed differs from rocket to rocket, it would be nice if there was some way to control how aggressively it turns inside the program.

[rebelcommando1807]

This is just something that I think would be nice to see.. Any chance you can detect if an engine has flamed out or is otherwise non-functional, and have the program detect the change in MaxThrust accordingly? Would make quite a few rockets easier to program for, certainly.. And planes, for that matter.

[JewelShisen]

I'll take another look at the steering.

The reason it changed is because when a vessel is in orbit, it was wildly oversteering causing wobbles that last forever. It also did rotations extremely slowly.

I may not have tested it with enough variety of rocket designs. If I can't get it fixed I'll revert back to the old method.

Pitching over during a gravity turn is usually a sign of not enough torque but it could also be understeering on the part of the mod. Would you be will to send me a private message with your craft file attached?

message sent with craft and program.

[Dunbaratu]

In the code I've split vectors and rotations back into separate types as part of exposing :X :Y and :Z.

I had added an implicit cast that I thought would have automatically converted vectors back into rotations when used in expressions. It looks like I just need to go back over it and make it more explicit.

That's good. Also, like I said I'd have no problem doing it myself if I had access to arcsin, arccos, and arctan.

[Camacha]

This thread is making it painfully clear that my high school math has pretty much gone down the drain. Time to do some catching up, I am afraid. At the same time this is why I adore this game; it makes math and physics less theoretical and a lot more accessible. If only something like this existed when I was in school...

(If anyone has some decent resources/websites on math I would - of course - be glad to hear about them.)

[SiliconPyro]

This thread is making it painfully clear that my high school math has pretty much gone down the drain. Time to do some catching up, I am afraid. At the same time this is why I adore this game; it makes math and physics less theoretical and a lot more accessible. If only something like this existed when I was in school... (If anyone has some decent resources/websites on math I would - of course - be glad to hear about them.)

Maybe check out khanacademy.com

[check]

Can someone explain what exactly surface velocity is measuring? Because it's not measuring my velocity with respect to the ground. That's for sure.

[DoctorCruz]

When I downloaded and tried to install, I got a texture error. I then tried to add the textures to the specified path and that did not work either. How do I fix it?

[bazilshep_]

Surfacevelocity appears to be measuring the horizontal components of the surface velocity.

[Camacha]

I'll take another look at the steering.

The reason it changed is because when a vessel is in orbit, it was wildly oversteering causing wobbles that last forever. It also did rotations extremely slowly.

I may not have tested it with enough variety of rocket designs. If I can't get it fixed I'll revert back to the old method.

I kind of feel that kOS might be taking too many decisions on its own here. In my mind, the goal is to be able to write your own autopiloting system. If you shove too much decision making behind the scenes to keep it accessible, I feel one might just use MechJeb instead.

I'd love to be able to write a program that checks and rechecks attitude, instead of just ordering a vector and kOS - succesfully or not - working it all out by itself.

Maybe check out khanacademy.com

I kind of stumbled across this and remembered a friend recommending it yea long ago. Pretty neat stuff. Thanks!

[Payload]

I kind of feel that kOS might be taking too many decisions on its own here. In my mind, the goal is to be able to write your own autopiloting system. If you shove too much decision making behind the scenes to keep it accessible, I feel one might just use MechJeb instead.

I'd love to be able to write a program that checks and rechecks attitude, instead of just ordering a vector and kOS - succesfully or not - working it all out by itself.

I kind of stumbled across this and remembered a friend recommending it yea long ago. Pretty neat stuff. Thanks!

The problem is speed. It just isn't fast enough to do that and everything else it needs to do. Your steering would be all over the place.

[Camacha]

The problem is speed. It just isn't fast enough to do that and everything else it needs to do. Your steering would be all over the place.

Fair point. What is causing this? Is this some inherent restriction of KSP, an artificial restriction to keep kOS from using too much processor time or something else?

[DoctorCruz]

I am trying to install this add-on but I get a missing textures error. I dropped the contents of the kOS folder in the main KerbalSpaceProgram folder. That added the controller part but not the textures for the console. The error message said it is looking for the textures in the Kerbal Space Program\Plugins\PluginData\kOS\gfx path so I put them there, but it still doesn't work. What am I doing wrong?

[Sma]

Wow. I guess now that I think of it I never actually tried driving straight through the rocks. I just assumed they were real physical barriers and took care to never run into them.

As I recall I think I avoided them too, until I saw Scott Manley drive through them. That or maybe it was once when I was going "oh no....oh NO..NO.....oh....uhh ok that happened...holographic rocks".

I'm having issues with it pitching over for the gravity turn on my test rocket.

Yeah, I'm having a similar problem, except it's before the gravity turn. Seems like the same problem, though less pronounced, as when I was trying to use V(?,?,?) the other day. Sometimes it seems to be more wobbly than others, and sometimes it completely spins around and then corrects itself just before reaching the altitude where i want the gravity turn to start. Once it's on the 45 it's good, though it does wobble a few degrees here and there. Also, it doesn't seem to do it until maybe 1000m give or take, possibly due to fuel movement. It's almost as if it's not fully locking in the coordinates, or similar to launching without SAS turned on.

I have tried adding/removing SAS units, including keeping an ASAS on the top of the rocket, and adding a SAS unit to the bottom of the rocket.

I'm using this to set the orientation before launch, and in a loop after launch, if altitude is below 10000:

lock steering to heading 90 by 90.

Maybe I shouldn't keep setting it in the loop? Though I think I had that originally and it didn't seem to make a difference. Oh, also I have tried having it set timewarp, sometimes it'll let physical warp happen, other times it won't, but when it does the wobbling/pitching over is much more pronounced (like usual I guess).

I think I'll try setting steering to up next, and leave it at that until the gravity turn.

Update: So I had a strange thing happen, now all of a sudden the launch clamps disappear, and the whole rocket shifts sideways towards the east (while still pointing up) as if it was teleported. A few launches like this resulted in the script module crashing into the water pipes, and rapid unplanned disassembly ensued. I thought this was because of the game running most of the day (mostly just sitting idle) so I restarted it. First launch (with locking steering to up) went mostly ok, though it still drifted. Next launch though (using lock steering to up + R(0,0,180) ) the teleport happened again, but the rocket did seem to stay relatively straight this time, with maybe half a degree wobble/going off course. I'll try some other things and see if I can stop it.

Edited September 16, 2013 by Sma

[Dunbaratu]

Surfacevelocity appears to be measuring the horizontal components of the surface velocity.

That's correct. If you want the *actual* surface speed like the navball gives you have to use the velocity:surface vector's components and run them through the Pythagorean formula like so:

set velSurf to velocity:surface.

set realSurfSpeed to ( velSurf:x^2 + velSurf:y^2 + velSurf^z^2 ) ^ 0.5 .

The reason for the intermediate step of "set velSurf to velocity:surface" is that the system doesn't understand a term with two colons in it like so: "surface:surface:x".

[henryrasia]

Sorry if this is common knowledge, but what is the difference between volumes and files? And how can you create and edit a file within a volume? Because so far I've been writing on the volume itself it seems.

Also, would anyone be interested on doing a full on tutorial on kOS, written or in video form? Thanks.

[check]

That's correct. If you want the *actual* surface speed like the navball gives you have to use the velocity:surface vector's components and run them through the Pythagorean formula like so:

set velSurf to velocity:surface.

set realSurfSpeed to ( velSurf:x^2 + velSurf:y^2 + velSurf^z^2 ) ^ 0.5 .

The reason for the intermediate step of "set velSurf to velocity:surface" is that the system doesn't understand a term with two colons in it like so: "surface:surface:x".

See, I was having a weird issue where the (x,y,z) components were showing large values >1 or <-1 and my ship was stationary. I decided to check the surface velocity on just a command pod while on the ground and then again while on top of a 60m tall launch clamp. On the ground I was getting near 0 but on the clamp I was showing a steady value of ~1.2 m/s in th x component, with the other two showing near 0. So I thought something else was going on. But after resetting the game and trying it again, the numbers look ok now. Weird...

[Dunbaratu]

See, I was having a weird issue where the (x,y,z) components were showing large values >1 or <-1 and my ship was stationary. I decided to check the surface velocity on just a command pod while on the ground and then again while on top of a 60m tall launch clamp. On the ground I was getting near 0 but on the clamp I was showing a steady value of ~1.2 m/s in th x component, with the other two showing near 0. So I thought something else was going on. But after resetting the game and trying it again, the numbers look ok now. Weird...

I've often had wonky behavior from the KOS unit that could be fixed with the far less drastic step of rebooting the KOS unit rather than restarting the whole game. (rightclick the part, toggle power off, then toggle power on again) Give that a try first next time there's a problem. Sometimes it works, sometimes it doesn't. I've found it especially useful when its one of the problems that causes it to start deciding to make the text you type invisible.

[Canadientire]

Its a cool plugin, but in the next update is it possible to make it compatible with the RemoteTech plugins? As it doesn't work with them: When you press spacebar, it stages even if you're writting, and the brake, lights, RCS, SAS, action groups and abort are activated to when their button is pressed.

[Secret Squirrel]

Sorry if this is common knowledge, but what is the difference between volumes and files? And how can you create and edit a file within a volume? Because so far I've been writing on the volume itself it seems.

Also, would anyone be interested on doing a full on tutorial on kOS, written or in video form? Thanks.

On the front page of this thread are links to videos to help get started and explain new features. They are not too in depth, but if you look at some peoples codes on here you can get an idea of how it works. Also trial and error seems to work for me !

[Dunbaratu]

Sorry if this is common knowledge, but what is the difference between volumes and files? And how can you create and edit a file within a volume? Because so far I've been writing on the volume itself it seems.

Also, would anyone be interested on doing a full on tutorial on kOS, written or in video form? Thanks.

I was confused enough about this when I started that once I finally did work it out from trial and error I decided it needed to be documented better so I wrote this page on the wiki: http://kos.wikia.com/wiki/Archive_and_Volumes

I hope that helps.

[Sma]

Sorry if this is common knowledge, but what is the difference between volumes and files? And how can you create and edit a file within a volume? Because so far I've been writing on the volume itself it seems.

Also, would anyone be interested on doing a full on tutorial on kOS, written or in video form? Thanks.

Lets see if this helps explain it, at least as far as I know anyway ... A volume is similar to a music cd (people still know what those are right? lol), where the volume is the cd, and files are the songs on that volume. In the case of KSP/kOS, each rocket that has the kOS module is a separate volume (don't know if they're accessible from other rockets), if you notice when you right click on the kOS module, it has an ID. the archive volume is based in KSC so you have to be in range of KSC. Basically the idea is you can load up your archive with various files that you'd reuse, like say a launch/orbit program, a docking program, a rover program etc. You copy the programs you need from the archive to the current volume/rocket so you don't have to worry about being in range of KSC. This way you can have different programs loaded up on different rockets. If you leave that rocket, or quick save those programs are saved in the persistent/quick save file I believe.

Also, I don't believe you can edit programs that are in the rockets volume outside of KSP/kOS, only the ones in the archive volume. So you can use notepad, or the kOS IDE someone has created (that does syntax highlighting based on kerboscript), to edit programs outside of KSP and then copy them to the rocket, or like I do, I just "switch to archive" and run them from there. Though once I get back into actually playing KSP and not just testing things (whenever RT2 comes out of play testing) then I'll just put my programs in archive and copy them to the rocket(s) as needed.

BTW, the archive folder, at least for now, is located in Kerbal Space Program\Plugins\PluginData\Archive