Parts & Vessels Undocking Radial Docking Ports Can Cause Navball to Become Unresponsive and Timewarp to Freeze Craft in Place.

[Anth]

KSP2 Version: 0.1.0.0.20892
KSP2 Version: 0.1.1.0.21572
KSP2 Version: 0.1.2.0.22258
Operating System: Windows 10
CPU+GPU+RAM: i9 9900K+3070ti+32GB
Mods Used: None

Three issues are happening here:

1. The control point on undocking a craft can in some cases not reset to the undocking craft.
2. The Navball freezes and doesnt reflect the correct orientation of the craft but Control From Here change it to the current orientation like a photo.
3. On using Timewarp and then going back to normal time the craft freezes in place even if the engines are activated.

 

Below are Two Different Situations Using the Same Crafts to Make this Happen for Different Docking Ports:

1. Tug Undock Bug Test:

Spoiler

Video Evidence of How I Make the Tug do it.:
https://youtu.be/L116ZKrK8GY

Steps to Replicate:

1. Load 2ndPreTugPRadTanktoRadTankDock
2. Reload 2ndPreTugPRadTanktoRadTankDock (corrects speed and other possible issues) if first loaded from Main menu
3. Wait for docking.
4. Undock tug's docking port.
5. Try pitching up which it will spin instead
6. Select control from here on the tugs docking port and pitch up
7. Timewarp for a moment and return to normal time.
8. Go to 100% throttle.

Save File For Testing From Video:
https://www.dropbox.com/s/fwvpzovt7t0z9hl/2ndPreTugPRadTanktoRadTankDock.zip?dl=0
Save File For Testing End Result:
https://www.dropbox.com/s/6mg3gvuyzrq11b1/3rdTugPRadTanktoRadTankDock.zip?dl=0

Log Files:
https://www.dropbox.com/s/f14idu1273tigd4/UndockingTugLog.zip?dl=0

Craft File:
https://www.dropbox.com/s/yxa33wbt9podt4m/0120Tank4xRadialPortsProbe2.zip?dl=0

In this case what is happening is:

• Bug will occur when undocking the tugs docking port
• Doing it to the tank on the other side it will function correctly.

2. Tank Undock Test:

Spoiler

Video Evidence of How I Make the Tank do it:
https://youtu.be/i9iJWFRwMok

Steps to Replicate:

1. Load 2ndPreTugPRadTanktoRadTankDock
2. Reload 2ndPreTugPRadTanktoRadTankDock (corrects speed) if first loaded from Main menu
3. Undock Using the docking port on the tank with radial docking ports.
4. Wait for dock.
5. Change to tug.
6. RCS forward the tug a little bit to get it to dock.
7. F5 and F9 to counter battery of end tank not being charged by solar panels.
8. Undock end tank using its docking port.
9. Turn on SAS and Click on Anti normal (Anything that is far from its orientation.)
10. Wait for some movement and use timewarp.
11. Turn off timewarp and note stuck position.

Save File For Testing From Video:
https://www.dropbox.com/s/fwvpzovt7t0z9hl/2ndPreTugPRadTanktoRadTankDock.zip?dl=0
Save File For Testing End Result:
https://www.dropbox.com/s/ufle5ghkinc3oz4/3rdRadTanktoRadTankTugLast.zip?dl=0

Log Files:
https://www.dropbox.com/s/5jhthquhlr9sgay/UndockingTankLog.zip?dl=0

Craft File:
https://www.dropbox.com/s/yxa33wbt9podt4m/0120Tank4xRadialPortsProbe2.zip?dl=0

In this case what is happening is:

• Bug will occur when undocking the outside Tanks docking port
• Doing it to the Tug's docking port will function correctly.
• Note undocked craft doesnt have propulsion like the previous test.

Additional Test Which I Did Earlier in Testing That I Only Have a Save File For:

Tank Radial and Node Undock Test:

Spoiler

Screenshot of the Craft:
https://i.imgur.com/vyoG1t5.png

Step to Replicate:

1. Load RadialDockandNodeDockTestSave
2. Undock using the docking port which is connected to the Probe Core that is connected to the MultiPointConnector.
3. Turn on SAS and Click on Anti normal (Anything that is far from its orientation.)
4. Wait for some movement and use timewarp.
5. Turn off timewarp and note stuck position.

Save File:
https://www.dropbox.com/s/nuaspjj29yz751y/RadialDockandNodeDockTestSave.zip?dl=0

Log Files:
https://www.dropbox.com/s/i2dquuccx3ng23j/RadialNodeLog.zip?dl=0

Important Information:

• Log Files indicate the following after disengaging timewarp: [Flight] PartComponent must have a root part!

Additional Information:

• Testing in 0.1.0.0 Appears to have the bugged craft move in a radial out direction even with no engines being fired when there is no directional change in 0.1.2.0
• Testing in 0.1.1.0 Appears to be the same as 0.1.2.0

Work Arounds:

• Undock from the other docking port and then change to the craft
• F5, F9 (Quick Save, Load Quick Save)

Control Point Not Resetting Aspect is Similar to the Following:

 

Edited June 1, 2023 by Anth12

[schlosrat]

I'm getting a similar (possibly same) issue. I don't have the pre-issue saves, but I could probably reconstruct them.

I've two identical small craft (23 parts) in LKO. Both have a single Clamp-o-tron Jr on the top of their Mk1 command pod. Both craft were perfectly controllable for attitude prior to docking, but in the save I've got now where they are undocked one is controllable with SAS and the other is not. Any attempt to control attitude with SAS for the uncontrollable craft (other than just Stability On) results in yaw & pitch causing the craft to spin about it's CM.  Unlike one of the videos you posted I've not noticed a lot of roll, just crazy yaw and pitch.

This happens when you use SAS to point to target, anti-target, node, prograde, normal, etc. All SAS modes except Stability On (padlock icon). The identical controllable craft survived undocking and works fine. With the uncontrollable craft the only thing I've found there are two things that will restore the ability to use SAS to point in a particular direction.

The first is to stage and separate the Mk1 pod from the tanks, SRW-125 and engine. If I stage the little guy then the pod is controllable as normal, but the debris is not as it has no control pod. In other words, after staging things seem normal.

These craft have two sources for torque, the smaller reaction wheels in the Mk1 pod, and an SRW-125 that's in the stage with the tanks and engine. If I disable torque on the Mk1 pod, the spinning problem persists, but if I disable torque on the SWR-125 that corrects the issue, though pointing is then more sluggish. I can leave torque enabled on the SRW-125 if I switch to a torque mode of Manual Only. Selecting All or SAS Only with torque enabled on the SWR-125 will make the craft spin as described above.

I've tried selecting different control points (the docking port or the Mk1 pod), but that had no effect. Only disabling SAS torque on the SWR-125 made a difference.

@Anth12, If you have multiple torque sources you might try disabling them one at a time to see which one is messed up. Of course we may have two different problems that just appear to have a similar effect. 

[Anth]

3 hours ago, schlosrat said:

I'm getting a similar (possibly same) issue. I don't have the pre-issue saves, but I could probably reconstruct them.

I've two identical small craft (23 parts) in LKO. Both have a single Clamp-o-tron Jr on the top of their Mk1 command pod. Both craft were perfectly controllable for attitude prior to docking, but in the save I've got now where they are undocked one is controllable with SAS and the other is not. Any attempt to control attitude with SAS for the uncontrollable craft (other than just Stability On) results in yaw & pitch causing the craft to spin about it's CM.  Unlike one of the videos you posted I've not noticed a lot of roll, just crazy yaw and pitch.

This happens when you use SAS to point to target, anti-target, node, prograde, normal, etc. All SAS modes except Stability On (padlock icon). The identical controllable craft survived undocking and works fine. With the uncontrollable craft the only thing I've found there are two things that will restore the ability to use SAS to point in a particular direction.

The first is to stage and separate the Mk1 pod from the tanks, SRW-125 and engine. If I stage the little guy then the pod is controllable as normal, but the debris is not as it has no control pod. In other words, after staging things seem normal.

These craft have two sources for torque, the smaller reaction wheels in the Mk1 pod, and an SRW-125 that's in the stage with the tanks and engine. If I disable torque on the Mk1 pod, the spinning problem persists, but if I disable torque on the SWR-125 that corrects the issue, though pointing is then more sluggish. I can leave torque enabled on the SRW-125 if I switch to a torque mode of Manual Only. Selecting All or SAS Only with torque enabled on the SWR-125 will make the craft spin as described above.

I've tried selecting different control points (the docking port or the Mk1 pod), but that had no effect. Only disabling SAS torque on the SWR-125 made a difference.

@Anth12, If you have multiple torque sources you might try disabling them one at a time to see which one is messed up. Of course we may have two different problems that just appear to have a similar effect. 

Not the same issue (not torque related that is)  though in my testing for this bug report I did have a craft start spinning out of control. Not sure what triggered it because it only happened a few times.

[schlosrat]

I think it may be related, actually. As you observed you tried to pitch up and it rolled instead. From what I've seen the thing I'm observing seems to be that reaction wheel control systems are getting messed up when undocking. This could account for why a pitch command gave you roll instead.

Have you tested using both manual control and SAS to effect the attitude? I suspect it won't matter which you use as both are just sending commands to the reaction wheels, which is where I think the problem is.

Further inspection in my test case reveals it's not that one vessel gets a messed up SWR-125 and the other doesn't. It's one vessel gets it's SWR-125 messed up and the other gets the Mk1 pod's reaction wheels messed up! And it's not just happening for one particular pair of vessels. I hadn't noticed that before as the one with the messed up Mk1 pod has sufficient control authority from it's working SWR-125 that it only wobbles a bit rather than spinning wildly when under SAS control.

Also, it's not something that's particular to a single pair of docked vessels. I've tested this with a similar, but different pair and they did the same thing. When you pick a docking port and decouple, it may be that the vessel you've got control of (the one who's docking port did the decoupling) is the one with the messed up SWR-125, or you may get the one with the messed up Mk1 pod. Regardless, both vessels get messed up when undocking in the test I've run.

I going to try this in a fresh campaign to be certain it's not something else pre-existing in the save, but it's not just a one-off with a particular pair of vessels, and it's happening for me in a completely clean and unmodified game - no mods are doing this.

I'll also try downloading your save to see if it does the same in my game.

Edited May 13, 2023 by schlosrat
added more details

[schlosrat]

Update. Using your save file I've confirmed that when undocking from the tug side the nav ball freezes and pitch commands give roll. Also, I've observed what appears to be the same general effect, though both the SWR-125 and the RC-001S on the tug seem to be messed up after it decouples from the containers. There is a messed up vector or vectors in the resulting decoupled craft so that it can't be controlled normally, and all SAS modes result in some sort of tumble. You can swap between the docking port and the probe core for "control from here" and get tumble directions to reverse, but tumble you will.

Decoupling from the radial port side and switching to the tug had a slightly different effect in that you could point to "prograde" and be stable, but the so-called "prograde" vector was only a few degrees off from radial minus, so in your save as well both resulting vessels are adversely affected after decoupling.

This was done in a brand new campaign, and using your save file, so it's not a problem local to either of our setups.

Edited May 13, 2023 by schlosrat