[1.9-1.10] Throttle Controlled Avionics

[allista]

28 minutes ago, kerbalfreak said:

I simply love this mod!

I've made a ion hover for Minmus, but I have some questions. Here a pic of it:

The central engine has low thrust, but high Isp. The other 4 has higher thrust and lower Isp.

Is possible to set it to only start using the 4 higher thrust engines when the central reaches 100% throttle? I set the central one to "unbalanced thrust", and the other 4 to "maneuver only", it works, but isn't exactly what i'm looking for.

Another thing, it seems very conservative when cruising or stopping, taking much more time than needed. The throttle only stays at 50% and it inclines much less than it could. I changed some configs in the profiles, but didn't made much difference. Is something else I could try? Now I'm taking full manual control, or setting MechJeb to land when I want to stop.

Thanks for the awesome mod!

You can make two engine profiles: one for takeoff/landing where you set the four to thrust & maneuver, and the other for flight, where, as you do, you set them to maneuver only. You'll have to switch between them manually, but the just a single click.

As for the speed, I'm not sure. You say "central reaches 100%" - does it mean the engine has non-zero thrust reaction time? If so, considering that the other four are for maneuvering, TCA will tilt slowly and only to a degree, so as not to loos TWR. Again, try to use the four as thrust & maneuver.

Another reason TCA takes it slow may be the calculation of brake path: if you have low TWM, it would take much time to accelerate/decelerate, especially if the only thrust provider is a slow-response engine.

[kerbalfreak]

2 hours ago, allista said:

You can make two engine profiles: one for takeoff/landing where you set the four to thrust & maneuver, and the other for flight, where, as you do, you set them to maneuver only. You'll have to switch between them manually, but the just a single click.

As for the speed, I'm not sure. You say "central reaches 100%" - does it mean the engine has non-zero thrust reaction time? If so, considering that the other four are for maneuvering, TCA will tilt slowly and only to a degree, so as not to loos TWR. Again, try to use the four as thrust & maneuver.

Another reason TCA takes it slow may be the calculation of brake path: if you have low TWM, it would take much time to accelerate/decelerate, especially if the only thrust provider is a slow-response engine.

Thanks, I will try this! Using the two profiles will work much better than setting action groups like I was doing.

Looking in the config file, the central one seems to have a response time of 0.5, the other 4 are stock dawns. I was saying to only start using the four when it needed more than 100% of the thrust of the central one (the "trust limiter", not the time to respond and reach the desired throttle). Like this, when the need for thrust is increasing:

Central: 55% - Other four: 0%
\/
Central: 80% - Other four: 0%
\/
Central: 100% - Other four: 30%
\/
Central: 100% - Other four: 70%

I'd like to use all in the thrust and maneuver mode, with this behavior. But the profiles will do fine.

The TWR of the vessel is around 4-5 in Minmus. It's not much, but TCA takes forever to gain speed or stop, much more than in manual mode or with MechJeb2. Seems like it ignores the (over)power of the reaction wheels for turning the vessel.

Some pics. With TCA, barely reducing the speed:



MechJeb2 killing the horizontal speed while maintaining altitude before landing. It can stop the vessel relatively fast.

Pointing retrograde also works, but it is a pain to control vertical speed and altitude by hand.

There isn't a why to make TCA ignore all the safety protocols and go for the fastest possible way?  With the gravity and lack of atmosphere in Minmus, and the reaction wheels, it will be no problem.

Thanks again for the awesome mod!

Edited November 16, 2017 by kerbalfreak

[allista]

1 hour ago, kerbalfreak said:

Thanks, I will try this! Using the two profiles will work much better than setting action groups like I was doing.

Looking in the config file, the central one seems to have a response time of 0.5, the other 4 are stock dawns. I was saying to only start using the four when it needed more than 100% of the thrust of the central one (the "trust limiter", not the time to respond and reach the desired throttle). Like this, when the need for thrust is increasing:

Central: 55% - Other four: 0%
\/
Central: 80% - Other four: 0%
\/
Central: 100% - Other four: 30%
\/
Central: 100% - Other four: 70%

I'd like to use all in the thrust and maneuver mode, with this behavior. But the profiles will do fine.The TWR of the vessel is around 4-5 in Minmus. It's not much, but TCA takes forever to gain speed or stop, much more than in manual mode or with MechJeb2. Seems like it ignores the (over)power of the reaction wheels for turning the vessel.

Some pics. With TCA, barely reducing the speed:

MechJeb2 killing the horizontal speed while maintaining altitude before landing. It can stop the vessel relatively fast.

Pointing retrograde also works, but it is a pain to control vertical speed and altitude by hand.
There isn't a why to make TCA ignore all the safety protocols and go for the fastest possible way?  With the gravity and lack of atmosphere in Minmus, and the reaction wheels, it will be no problem.Thanks again for the awesome mod!

Such differential thrust as you describe is not possible with TCA yet. But I'll explore the possibility.

***

And that's what normal operation in low-G looks like when you use instant-response engines:

https://www.twitch.tv/videos/201778656

Notice the constant acceleration pattern: the craft accelerates until about half the distance, then decelerates slowly the other half. This way it's easy to control overshooting and it's faster than accelerate-cruise-decelerate (which also could be achieved using maximum speed limit). Also notice that in the beginning TCA uses most of the thrust to accelerate, tilting greatly. But in the end it controls the deceleration much softer, so it doesn't need to tilt so much. So it's not that TCA can't use thrust to decelerate quickly; it just uses another strategy for VTOL navigation.

 

[agrock]

I noticed there were some throughout changes in latest release. Would it be okay to use it on KSP 1.3 (its certified for 1.3.1)?

[allista]

25 minutes ago, agrock said:

I noticed there were some throughout changes in latest release. Would it be okay to use it on KSP 1.3 (its certified for 1.3.1)?

Nope. The API of KSP 1.3.1 is not compatible with that if 1.3.0.

[agrock]

1 hour ago, allista said:

Nope. The API of KSP 1.3.1 is not compatible with that if 1.3.0.

Got  it. I ask because I am using one or two mods that were released only for 1.3.1 but work just the same.

[allista]

1 hour ago, agrock said:

Got  it. I ask because I am using one or two mods that were released only for 1.3.1 but work just the same.

Alas, not this one. Some methods in CelestialBody now return coordinates in another reference frame than in 1.3.0. So you can run TCA-3.5 on KSP 1.3, but orbital autopilots would be broken.

[kerbalfreak]

On 11/16/2017 at 7:04 PM, allista said:

Such differential thrust as you describe is not possible with TCA yet. But I'll explore the possibility.

***

And that's what normal operation in low-G looks like when you use instant-response engines:

https://www.twitch.tv/videos/201778656

Notice the constant acceleration pattern: the craft accelerates until about half the distance, then decelerates slowly the other half. This way it's easy to control overshooting and it's faster than accelerate-cruise-decelerate (which also could be achieved using maximum speed limit). Also notice that in the beginning TCA uses most of the thrust to accelerate, tilting greatly. But in the end it controls the deceleration much softer, so it doesn't need to tilt so much. So it's not that TCA can't use thrust to decelerate quickly; it just uses another strategy for VTOL navigation.

 

Loved that you replicated my hover! haha

I still think it goes slower than it could, both for accelerating to cruise or stopping. I think I will do this manually. Made a new one with more TWR, it's very easy. The hard/boring part is to keep it level and to hold altitude, which TCA do perfectly. When I want to fine tune, I use the waypoint function. To cover large distances the "jump to" is fantastic. It's working very well, no complaints. Doing this vehicle would be extremely hard and boring without your mod.

Loved it so much that the little hover gained a big and "kerbed" version 

[0something0]

On 11/12/2017 at 10:58 PM, 0something0 said:

I installed TCA into my save yesterday and it was working fine, even though the UI was all messed up, but the features I needed at the moment, flying VTOLs, was working correctly. However, when I reloaded the game, the Hover feature simply did not do anything. I tested the mod with only the necessary dependencies and it worked though. And no, I don't know where the logs are: the game crashed due to an unrelated reason, probably due to a lack of memory on my potato, but I'm pretty sure it only has the logs leading up to the moment of the crash. 

https://imgur.com/a/jq1jq broken(?) UI

https://imgur.com/a/IV6qU Mod folder

I couldn't find time during the week to test this out some more, but I have now. Here is what I found. 

-Stock engines on modded save works fine

-The craft that I was experiencing issues with, which relied on spitfire mod engines, works albit very slowly, especially at takeoff.

[Steel Dragon]

How do you add TCA compatibility to a set of engines that TCA does not detect naturally. Particularly I would love to get TCA to control the fan engines in this mod
 

 

[allista]

41 minutes ago, Steel Dragon said:

How do you add TCA compatibility to a set of engines that TCA does not detect naturally. Particularly I would love to get TCA to control the fan engines in this mod
 

 

TCA works with the stock ModuleEngines and the modules that directly inherit from it. If it doesn't see an engine at all, this means the module that actually generates thrust is implemented in a different fashion.

Unfortunately, in many cases TCA uses ModuleEngines directly, without wrapping it into an abstraction layer. So currently it's not even possible to write a "driver" for a custom thrust provider.

If this request gains enough support I would explore the idea of adding the "engine driver" API for other modders to use to add TCA compatibility to custom modules; but I already can tell that in terms of work it would cost a new major release.

[jinks]

I'm having a bit of trouble trying to get TCA to launch my ships to orbit...

 

First: Do I really have to micromanage every single ships launch configuration? I have to create a separate profile for every single stage of a multistage launch or TCA just turns engines and spent-stage retroboosters on more or less randomly.

 

Second: No matter what I do, TCA always launches ships at a very steep angle and wastes tons of dV on getting up to orbital speed.

Up until now I have used this kOS script for launches (GitHub), but well, it's kOS, so quite a resource hog (it takes about 7-8 seconds alone to compile an start on launch) and it means I have to include and manage kOS parts on every single ship.

But that script gets me into orbit for several hundred m/s of dV less than TCA.

 

As an example, here's the trajectory of the same ship right after it turned off the engines to coast to AP.

Once launched with TCA (steepness 10%)

and once with the kOS script

Some observations:

• the kOS script turns much more aggressively early on, going to a 50° trajectory at about 5km up
• the KOS script keeps the nose on prograde instead of on the circularization node while coasting to minimize drag
• TCA constantly complains about low control authority. - My ships rely on control surface fins for atmo steering, does TCA not recognize those?

What can I do to make TCA behave more like the linked script? And can you maybe have a look at the math in that script to improve TCA's gravity turns? (It relies on an exponential curve which has, so far, worked for every ship I've thrown at it.)

As a side note: ToOrbit parameters revert to 71, 0.9, 50% on every launch. Is there a place to set different defaults?

 

Edited November 22, 2017 by jinks

[ss8913]

@allista I've run into a problem... the new version doesn't behave like the 1.3.0 one (the last release for 1.3.0) did, and in a cripplingly negative way.  Allow me to explain:

Most of my space program revolves around mk3 and mk4 spaceplanes using KSPIE nuclear engines, with VTOL capability.  This involves one or two horizontally-facing main engines, and 2 to 6 downward-pointing engines for VTOL thrust.  This worked perfectly with the aforementioned version of TCA.  I found that if I set the horizontal engines to manual control and the vertical engines to 'thrust and maneuver', and then set TCA to 'level' or 'stop', set the altitude, engaged the engines, and enabled TCA, it would lift to the desired altitude, and then I could push the 'Cruise' button, set the heading to 90 degrees and the horizontal speed to ~200 m/s, it'd use the horizontal engines, and do exactly what was asked/expected.  At which point I'd disengage the VTOL engines, disengage TCA, and fly to orbit on the main horizontally-facing engines.

OK, fast forward to today, 1.3.1, new TCA, same ship design.  I notice:

1. The altitude hold is always with reference to the ground, regardless of the 'follow terrain' button.  I cannot set an MSL altitude anymore
2. The altitude number used to be red when below ground, green when above ground, now it's always red.  May be linked to #1
3.  Horizontal engines always 0 thrust when set as above; if I set it to thrust & maneuver, it attempts to approximate the previous behavior, so that might be just a change in the new version which doesn't break functionality
4.  Now for the game-breaking bug - as soon as I push the CRUISE button, my frame rate gets cut in half or worse, instantly, and only a revert-to-launch/SPH will remedy the situation.  This is definitely new, and I would hope is a bug that can be fixed
5. Side note - "Smart Engines" still does not behave the way I'd expect (I thought smart engines would do what I'm describing above, but, no, it just prevents my horizontal engines from being activated, period.  If I activate them manually, it just shuts them off immediately.  But the previous version did that, too.)

In the logs, I see a lot of:

(Filename: C:/buildslave/unity/build/artifacts/generated/common/runtime/UnityEngineDebugBindings.gen.cpp Line: 42)

[ThrottleControlledAvionics: 17:54:07.599] P-2-B-V.AttitudeControl: compute_steering: Invalid argumetns:
current (0, 0, 0); |v| = 0
needed (0.3672802, -0.04080358, 0.929215); |v| = 1
current thrust (0, 0, 0); |v| = 0
 
(Filename: C:/buildslave/unity/build/artifacts/generated/common/runtime/UnityEngineDebugBindings.gen.cpp Line: 42)

[ThrottleControlledAvionics: 17:54:07.601] P-2-B-V.AttitudeControl: compute_steering: Invalid argumetns:
current (0, 0, 0); |v| = 0
needed (0.3673006, -0.04082808, 0.9292058); |v| = 1
current thrust (0, 0, 0); |v| = 0

 

[allista]

6 hours ago, ss8913 said:

@allista I've run into a problem... the new version doesn't behave like the 1.3.0 one (the last release for 1.3.0) did, and in a cripplingly negative way.  Allow me to explain:

Most of my space program revolves around mk3 and mk4 spaceplanes using KSPIE nuclear engines, with VTOL capability.  This involves one or two horizontally-facing main engines, and 2 to 6 downward-pointing engines for VTOL thrust.  This worked perfectly with the aforementioned version of TCA.  I found that if I set the horizontal engines to manual control and the vertical engines to 'thrust and maneuver', and then set TCA to 'level' or 'stop', set the altitude, engaged the engines, and enabled TCA, it would lift to the desired altitude, and then I could push the 'Cruise' button, set the heading to 90 degrees and the horizontal speed to ~200 m/s, it'd use the horizontal engines, and do exactly what was asked/expected.  At which point I'd disengage the VTOL engines, disengage TCA, and fly to orbit on the main horizontally-facing engines.

OK, fast forward to today, 1.3.1, new TCA, same ship design.  I notice:

1. The altitude hold is always with reference to the ground, regardless of the 'follow terrain' button.  I cannot set an MSL altitude anymore
2. The altitude number used to be red when below ground, green when above ground, now it's always red.  May be linked to #1
3.  Horizontal engines always 0 thrust when set as above; if I set it to thrust & maneuver, it attempts to approximate the previous behavior, so that might be just a change in the new version which doesn't break functionality
4.  Now for the game-breaking bug - as soon as I push the CRUISE button, my frame rate gets cut in half or worse, instantly, and only a revert-to-launch/SPH will remedy the situation.  This is definitely new, and I would hope is a bug that can be fixed
5. Side note - "Smart Engines" still does not behave the way I'd expect (I thought smart engines would do what I'm describing above, but, no, it just prevents my horizontal engines from being activated, period.  If I activate them manually, it just shuts them off immediately.  But the previous version did that, too.)

In the logs, I see a lot of:

(Filename: C:/buildslave/unity/build/artifacts/generated/common/runtime/UnityEngineDebugBindings.gen.cpp Line: 42)

[ThrottleControlledAvionics: 17:54:07.599] P-2-B-V.AttitudeControl: compute_steering: Invalid argumetns:
current (0, 0, 0); |v| = 0
needed (0.3672802, -0.04080358, 0.929215); |v| = 1
current thrust (0, 0, 0); |v| = 0
 
(Filename: C:/buildslave/unity/build/artifacts/generated/common/runtime/UnityEngineDebugBindings.gen.cpp Line: 42)

[ThrottleControlledAvionics: 17:54:07.601] P-2-B-V.AttitudeControl: compute_steering: Invalid argumetns:
current (0, 0, 0); |v| = 0
needed (0.3673006, -0.04082808, 0.9292058); |v| = 1
current thrust (0, 0, 0); |v| = 0

 

First of all, please, read the manual about Smart Engines. They're made for completely different thing.

Second, do you have any terrain enhancement mods installed? There were reports about them breaking TCA's ability to measure altitude.

Other problems may be derivatives of this altitude problem.

Edited November 23, 2017 by allista

[allista]

8 hours ago, jinks said:

I'm having a bit of trouble trying to get TCA to launch my ships to orbit...

 

First: Do I really have to micromanage every single ships launch configuration? I have to create a separate profile for every single stage of a multistage launch or TCA just turns engines and spent-stage retroboosters on more or less randomly.

 

Second: No matter what I do, TCA always launches ships at a very steep angle and wastes tons of dV on getting up to orbital speed.

Up until now I have used this kOS script for launches (GitHub), but well, it's kOS, so quite a resource hog (it takes about 7-8 seconds alone to compile an start on launch) and it means I have to include and manage kOS parts on every single ship.

But that script gets me into orbit for several hundred m/s of dV less than TCA.

 

As an example, here's the trajectory of the same ship right after it turned off the engines to coast to AP.

Once launched with TCA (steepness 10%)

and once with the kOS script

Some observations:

• the kOS script turns much more aggressively early on, going to a 50° trajectory at about 5km up
• the KOS script keeps the nose on prograde instead of on the circularization node while coasting to minimize drag
• TCA constantly complains about low control authority. - My ships rely on control surface fins for atmo steering, does TCA not recognize those?

What can I do to make TCA behave more like the linked script? And can you maybe have a look at the math in that script to improve TCA's gravity turns? (It relies on an exponential curve which has, so far, worked for every ship I've thrown at it.)

As a side note: ToOrbit parameters revert to 71, 0.9, 50% on every launch. Is there a place to set different defaults?

 

The low steepness should have made the trajectory much more elongated, so this is probably a bug. Thanks for the report.

Nose direction doesn't make much of a difference with stock aerodynamics (less so with FAR) above 50km; while missing apoapsis node may be fatal. But I'll see if that behavior could be improved.

Low control authority is not a calculated value, but a fact that TCA simply states. It means that using all your reaction control (including surfaces) it cannot decrease the attitude error to be less than 10 deg for a long enough time.

Yes, TCA is configured per ship, so you cannot set the defaults for autopilots. I can think of adding some sort of database for this case, but really, does setting of 3 parameters is so much of a burden? After all, KSP is anything but a strategic game, so the notion of micromanagement is hardly applicable.

Edited November 23, 2017 by allista

[ss8913]

3 hours ago, allista said:

For you all, please, read the manual about Smart Engines. They're made for completely different thing.

Second, do you have any terrain enhancement mods installed? There were reports about them breaking TCA's ability to measure altitude.

Other problems may be derivatives of this altitude problem.

I have Kopernicus installed, for Astronomer's Visual Pack... but... not sure that counts as a terrain mod?  I can try uninstalling that... the only issue I'm reporting that's a serious bug with no workaround that I can find is the framerate one, but yeah I can see how the framerate problem could be caused by a calculation going into some kind of loop... I'll go back to SVE and remove Kopernicus and try again. 

 

UPDATE:  Yes, this fixes it.  It seems to be an incompatibility between Kopernicus and TCA, which is odd because in KSP 1.3.0 I used these two mods together without any issue.  Removing AVP and Kopernicus (replaced with Stock Visual Enhancements) solves all of the problems 100%.

Edited November 23, 2017 by ss8913

[jinks]

9 hours ago, allista said:

The low steepness should have made the trajectory much more elongated, so this is probably a bug. Thanks for the report.

Might also be me doing something wrong... I tested with the stock CommSat LX, and TCA flipped it a few times when I set steepness too low.

9 hours ago, allista said:

Nose direction doesn't make much of a difference with stock aerodynamics (less so with FAR) above 50km; while missing apoapsis node may be fatal. But I'll see if that behavior could be improved.

kOS gets the AP to altitude when the ship is still only at 30,000m up in some cases so it matters a bit at least.

The script is relatively simple in its behaviour. Point to surface prograde until out of atmo, then point to node.

9 hours ago, allista said:

Low control authority is not a calculated value, but a fact that TCA simply states.

OK. I thought TCA might probably not try hard enough to turn because it thinks it can't. - Then the kOS script probably gets its advantage from leaning in very hard early on when the low speed doesn't pull you back to prograde so hard.

9 hours ago, allista said:

Yes, TCA is configured per ship, so you cannot set the defaults for autopilots. I can think of adding some sort of database for this case, but really, does setting of 3 parameters is so much of a burden?

Some of my heavy lifters have 6 or more stages. The stages are usually pretty simple, 1 big engine pointing out the back and 2-4 solids pointing forward to drop it back to the planet after separation.

What I see with TCA is that I have to create 6 profiles, set the active stage on each profile and then make sure only the backfacing engine for that stage is set to ON. - I'm currently in an "expansion" phase, so I launch a lot and I usually build custom launchers for each ship, maybe I just notice it more because of that.

Mechjeb or the linked script make the process a lot easier because they don't manage engines. They just stage and work with what's available.

This is obviously not as easy for TCA, but I find that I often only want the full capacity of TCA after the launch sequence is over. I don't really have a solution that won't cripple the main functionality of TCA and/or make it more tedious for the not-launch case.

A few random ideas:

• First time on the pad the ship is in PRELAUNCH state. As far as I know that state can never happen again later for a ship. Maybe use that as a trigger to turn off some management?
• While in ToOrbit mode (and on the first launch as determined by PRELAUNCH), ignore all stages except the current one, don't try to manage (turn on/off) engines, just throttle them as per usual.
• As soon as ToOrbit is done it can forget all about the above since it'll never happen again. That should make it at least somewhat simple code-wise. Two branches selected by a single external variable.

On another note, the landing sequence from TCA is the best thing I have ever seen. Neither Mechjeb nor even Trajectories come even close to the accuracy with which a TCA controlled ship puts me on-target. 

[ss8913]

On 11/23/2017 at 9:50 AM, jinks said:

Might also be me doing something wrong... I tested with the stock CommSat LX, and TCA flipped it a few times when I set steepness too low.

kOS gets the AP to altitude when the ship is still only at 30,000m up in some cases so it matters a bit at least.

The script is relatively simple in its behaviour. Point to surface prograde until out of atmo, then point to node.

OK. I thought TCA might probably not try hard enough to turn because it thinks it can't. - Then the kOS script probably gets its advantage from leaning in very hard early on when the low speed doesn't pull you back to prograde so hard.

Some of my heavy lifters have 6 or more stages. The stages are usually pretty simple, 1 big engine pointing out the back and 2-4 solids pointing forward to drop it back to the planet after separation.

What I see with TCA is that I have to create 6 profiles, set the active stage on each profile and then make sure only the backfacing engine for that stage is set to ON. - I'm currently in an "expansion" phase, so I launch a lot and I usually build custom launchers for each ship, maybe I just notice it more because of that.

Mechjeb or the linked script make the process a lot easier because they don't manage engines. They just stage and work with what's available.

This is obviously not as easy for TCA, but I find that I often only want the full capacity of TCA after the launch sequence is over. I don't really have a solution that won't cripple the main functionality of TCA and/or make it more tedious for the not-launch case.

A few random ideas:

• First time on the pad the ship is in PRELAUNCH state. As far as I know that state can never happen again later for a ship. Maybe use that as a trigger to turn off some management?
• While in ToOrbit mode (and on the first launch as determined by PRELAUNCH), ignore all stages except the current one, don't try to manage (turn on/off) engines, just throttle them as per usual.
• As soon as ToOrbit is done it can forget all about the above since it'll never happen again. That should make it at least somewhat simple code-wise. Two branches selected by a single external variable.

On another note, the landing sequence from TCA is the best thing I have ever seen. Neither Mechjeb nor even Trajectories come even close to the accuracy with which a TCA controlled ship puts me on-target. 

I'm still having issues with the landing sequence on graduated-response engines such as any of the nuclear engines in KSPIE.. and with the RCS contained within that same mod.  during the landing sequence especially, the TCA autopilot does not seem to apply enough control authority to execute the maneuvers fast enough, gets behind the curve, and crashes.  I can see that the problem is that it is actually not applying enough control input, based on the control position indicators.  This can be remediated by disabling RCS and using reaction wheels alone, *IF* the craft has enough RCS authority to get the job done... not sure why the KSPIE RCS causes this behavior, though.. maybe it's calculating the max power of the RCS thrusters and not the currently configured power levels, and is applying control authority proportionate to the max available?  that might be it..?

Further info - the KSPIE ATILLA engine, which can be used either as main thrust or maneuvering... does weird things with TCA.  ie, I can achieve a stable hover with them, but if I tell TCA to cruise at bearing X at speed Y... it doesn't ever input any pitch command to make that happen.  It does when using thermal engines or stock engines or ... other engines.. but it has some specific beef with anything marked as an attitude control module which comes from KSPIE.. both the normal RCS that it ships with plus these ATILLA thrusters.. no idea why it specifically has a problem with these specific engines, but.. if there's any tests you'd like me to run to narrow it down, I'm happy to.  Also, it's not like it *can't* achieve the desired pitch; if I tell TCA to "hold" and pitch manually, that works great.  Just the 'cruise' module doesn't seem to ever be applying the control input necessary.
 

Edited November 28, 2017 by ss8913

[ShangTai]

Hi,
I found a little problem, the Collapse Main Window do not work. When i try to use it i can see the TCA in the background but the main window will not hide. Nevertheless AutoShow option works fine. 
When i go to Tracking Station and then I back to vessel Main Window, will be collapsed.

[allista]

13 hours ago, ShangTai said:

Hi,
I found a little problem, the Collapse Main Window do not work. When i try to use it i can see the TCA in the background but the main window will not hide. Nevertheless AutoShow option works fine. 
When i go to Tracking Station and then I back to vessel Main Window, will be collapsed.

If you enable collapsing with auto-show, the window is only hidden when it's not hovered above by the mode cursor. So, naturally, when you press the up-arrow button, nothing happens until the cursor leaves the window.

[ShangTai]

10 hours ago, allista said:

If you enable collapsing with auto-show, the window is only hidden when it's not hovered above by the mode cursor. So, naturally, when you press the up-arrow button, nothing happens until the cursor leaves the window.

Auto-show works fine but, I wanna to hide it permanently with "up-arrow button" not with "auto-show". Up-arrow button, if i guess correctly, should hide window and only green or red TCA should stay. And when, i hover mouse near/above it, window should stay hidden.

[allista]

2 hours ago, ShangTai said:

Auto-show works fine but, I wanna to hide it permanently with "up-arrow button" not with "auto-show". Up-arrow button, if i guess correctly, should hide window and only green or red TCA should stay. And when, i hover mouse near/above it, window should stay hidden.

Yep, I see the problem now.

Thanks for the report. Will be fixed in the next release.

[danielboro]

i set target as mun base

then press lands

then press start

no landing yellow line tells me to chose waypoint or vessel

Filename: C:/buildslave/unity/build/artifacts/generated/common/runtime/UnityEngineDebugBindings.gen.cpp Line: 42)

File 'H:\‏‏Kerbal Space Program 1.3.1 havemod\GameData\000_AT_Utils\Plugins\..\000_AT_Utils.user' does not exist
 
(Filename: C:/buildslave/unity/build/artifacts/generated/common/runtime/UnityEngineDebugBindings.gen.cpp Line: 42)

ScaleModList: listSize 615 maxListSize 494
 
(Filename: C:/buildslave/unity/build/artifacts/generated/common/runtime/UnityEngineDebugBindings.gen.cpp Line: 42)

full log https://yadi.sk/d/gLwY_VAW3QG2VS

 

edit

even setting up a waypoint didn't help

cant land using tca autoland

 

edit 2

got it to start landing

from nave chose the waypoint on the base

then from orbit autopilot press the land

 

edit 3

my hotkey until now was `
but wen I try to set it I get "unebael to convert ` to keycode"

hope this can be fixed

 

Edited December 4, 2017 by danielboro

[ss8913]

Looks like you didn't install at_utils, which is required.   Do that. 

[danielboro]

20 hours ago, ss8913 said:

Looks like you didn't install at_utils, which is required.   Do that. 

I compered the zip from GitHub and gamedata
I have all the files (in dir 000_AT_Utils and dir ThrottleControlledAvionics)
non is missing