Jump to content

[1.12.x] Trajectories v2.4.5 (2023-08-22) : atmospheric predictions


Kobymaru

Recommended Posts

On 28.12.2017 at 9:12 AM, kerbalfreak said:

I'm only having problems with the predictions of this craft, the airbrakes slows it a lot in lower atmosphere, and the mod doesn't seem to take it into account, and I land much sooner. I enter in retrograde and have set it into the config.

Yes, airbrakes are currently broken in Trajectories, sorry about that. There is already a [GitHub issue](https://github.com/neuoy/KSPTrajectories/issues/58) about that and I will take a look.

Link to comment
Share on other sites

  • 2 weeks later...

Hi, great mod, thanks a lot. Been using it since a long time in stock KSP with stock aero. I especially love the KoS integration, reaally usefull for precision landing automation.

Since some time, I run RSS/RO/FAR. And unlike OP indicates, it does not seem to work correctly. I read in the previous comments that FAR is now known to be broken. Maybe an update in OP should indicate what Trajectories/FAR versions where the last to be working fine together? As I am on RSS/RO, I am of course looking for KSP 1.2.2 versions of these mods, was there a working version duo that worked on 1.2.2?

In any case, if any work is in progress for this issue, I'dd gladly help, at least to test it with RSS/RO/FAR mods

Thanks in any case for this great mod

Link to comment
Share on other sites

Future stages...ok

I have an idea, basically:

1) Quicksave the game

2) Decouple the stage

3) Save the profile data(?) on the new staged vehicle in a separate area/file ( a button in the PAW )

4) Reload quicksave and load the profile data ( a button in the PAW )

5) Until the profile is unloaded ( a button in the PAW ) trajectory is calculated based on the profile data gathered from the future stage

 

Maybe?

Edited by brusura
Link to comment
Share on other sites

On 12/30/2017 at 2:26 AM, Kobymaru said:

Yes, airbrakes are currently broken in Trajectories, sorry about that. There is already a [GitHub issue](https://github.com/neuoy/KSPTrajectories/issues/58) about that and I will take a look.

Regarding this, one thing I noticed the other night was other deployable items worked -- specifically, the inflatable heatshield and the petal fairings from BDB.  So taking a look their respective configs.

The airbrakes have this setting:

dragModelType = none

Whereas the other two are:

dragModelType = default

It makes me wonder how tricky it'd be to hack an alternate airbrake part that didn't have the control surface capability but did act more like a deployable radial heatshield.

Link to comment
Share on other sites

On 10.1.2018 at 11:29 AM, Garlik said:

Since some time, I run RSS/RO/FAR. And unlike OP indicates, it does not seem to work correctly. I read in the previous comments that FAR is now known to be broken. Maybe an update in OP should indicate what Trajectories/FAR versions where the last to be working fine together? As I am on RSS/RO, I am of course looking for KSP 1.2.2 versions of these mods, was there a working version duo that worked on 1.2.2?

I don't really know, since I didn't use FAR until very recently. I can try and do some testing.

The problem here is that if I understand correctly, RO is pretty much stuck with KSP 1.2.2 for the time being, but all fixes to Trajectories go into the 1.3.1 version of the mod. We will have to evaluate if a backport is possible with a reasonable workload.

Also, could you please give a bit more details on what "does not seem to work correctly" means?

 

18 hours ago, brusura said:

1) Quicksave the game

2) Decouple the stage

3) Save the profile data(?)

The last bit is the hard part and the question mark is well placed: currently the "profile data" is stored in KSP game objects that physically exist in the KSP universe, and we extract the drag information using KSP functions. I'm not entirely sure if it's even possible to separate this data into "ghost vessels", and I'm afraid that would be a rather large undertaking.

 

3 hours ago, sjbuggs said:

Regarding this, one thing I noticed the other night was other deployable items worked -- specifically, the inflatable heatshield and the petal fairings from BDB.  So taking a look their respective configs.

That's interesting, and it's a good start for debugging the issue, thanks! I do think we should handle this in the code instead of hacking an alternative airbrake part :)

Edited by Kobymaru
Link to comment
Share on other sites

7 hours ago, Kobymaru said:

Also, could you please give a bit more details on what "does not seem to work correctly" means?

Sure sorry, indeed not clear at all! Simply that the impact location is completely wrong. Retrograde correctly set in Trajectories, so as the craft itself.
example: projected impact before re-entry: close to KSC (East US Coast). Effective landing: in the ocean, west of the west US coast! So, largely off.
Of course, this is not a staging issue. During the atmospheric re-entry, keeping the craft retrograde, no fuel spent (thus no weight change), impact zone from Trajectories slowly moves and ends up thousands of kilometers off of its original landing prediction.

If you use RO/RSS, I suppose you should notice this too quite easily (if you do it on 1.2.2) with any craft. Others (reddit/discord) also note the same problem. I know a few people trying to get RSS/RO on 1.3.1, I'll ask them if the recent Trajectories version also show this bug on 1.3.1 and keep you informed

Thanks anyway. Great if it can be backported, but I'dd totally understand if not

Link to comment
Share on other sites

21 hours ago, Garlik said:

Sure sorry, indeed not clear at all! Simply that the impact location is completely wrong. Retrograde correctly set in Trajectories, so as the craft itself.
example: projected impact before re-entry: close to KSC (East US Coast). Effective landing: in the ocean, west of the west US coast! So, largely off.
Of course, this is not a staging issue. During the atmospheric re-entry, keeping the craft retrograde, no fuel spent (thus no weight change), impact zone from Trajectories slowly moves and ends up thousands of kilometers off of its original landing prediction.

OK, got it! That's the info I need.

21 hours ago, Garlik said:

If you use RO/RSS, I suppose you should notice this too quite easily (if you do it on 1.2.2) with any craft. Others (reddit/discord) also note the same problem. I know a few people trying to get RSS/RO on 1.3.1, I'll ask them if the recent Trajectories version also show this bug on 1.3.1 and keep you informed

Thanks but I doubt that it's gonna help much. I will do some testing myself. The only mods really relevant here should be FAR and RSS, I doubt that RO and it's dependencies are necessary for reproduction.

Link to comment
Share on other sites

Hey, @Kobymaru , I would like to report stuttering issue with Trajectories and FAR. It happens every ~2 seconds when I am descending on Kerbin in retrograde. I have latest KSP version, was well as FAR and Trajectories latest versions. And a ton of other mods, but I don't think any of them are related to this. While I'm descending, the log is giving off a message "Trajectories: Initializing cache" every ~2 seconds, which kind of coincides with the stuttering. That makes me suspect Trajectories is what's causing this.

Here's the log: https://ufile.io/gweez

If you need anything else to narrow down the issue, please let me know.

Link to comment
Share on other sites

On 21.1.2018 at 5:41 PM, aluc24 said:

Hey, @Kobymaru , I would like to report stuttering issue with Trajectories and FAR. It happens every ~2 seconds when I am descending on Kerbin in retrograde. I have latest KSP version, was well as FAR and Trajectories latest versions. And a ton of other mods, but I don't think any of them are related to this. While I'm descending, the log is giving off a message "Trajectories: Initializing cache" every ~2 seconds, which kind of coincides with the stuttering. That makes me suspect Trajectories is what's causing this.

Here's the log: https://ufile.io/gweez

If you need anything else to narrow down the issue, please let me know.

Thanks, I will take a look.

Link to comment
Share on other sites

On 22.1.2018 at 10:14 PM, aluc24 said:

Could you provide me with a quicksave file that I can load with only stock parts? Note that GitHub might prevent you from uploading the .sfs file directly, but you can rename it to .sfs.txt and it should be fine.

Link to comment
Share on other sites

On 1/30/2018 at 5:54 PM, Kobymaru said:

Could you provide me with a quicksave file that I can load with only stock parts? Note that GitHub might prevent you from uploading the .sfs file directly, but you can rename it to .sfs.txt and it should be fine.

I'm not sure anymore, but I think I might have been mistaken. Maybe these spikes are not due to Trajectories. I am now into my 4th day of hunting which mod causes these spikes, and I'm nowhere close to finding out. Logs now look clean (with Trajectories as well, no more "initializing cache" message), yet spikes are still there.
I will inform you if I find the cause, but I'm in the dark here.

Link to comment
Share on other sites

13 hours ago, aluc24 said:

I'm not sure anymore, but I think I might have been mistaken. Maybe these spikes are not due to Trajectories. I am now into my 4th day of hunting which mod causes these spikes, and I'm nowhere close to finding out. Logs now look clean (with Trajectories as well, no more "initializing cache" message), yet spikes are still there.
I will inform you if I find the cause, but I'm in the dark here.

Ok, do that. You could uninstall Trajectories and see if they still happen.

Link to comment
Share on other sites

On 2/2/2018 at 6:42 PM, dlrk said:

@aluc24

Try installing MemGraph and see if the red lines on the graph correlate with the stutters. If yes, then increasing the heap space with MemGraph will fix the stutters, provided RAM is available

I have MemGraph. No, the stutters do not correlate with red lines. Red lines happen every 30 seconds or so, while stuttering happens every 2-3 seconds

Link to comment
Share on other sites

On 5.2.2018 at 12:11 AM, Unieax said:

Hello,

I play RO now for quite a while and I found this nice mod today. I tried to add it via ckan but when I launch the game everything is as before without the mod installed aka nothing happened.

Do you know what I can do?

Do you happen to have blizzy's toolbar? I believe the old versions didn't show an icon unless you "created" it in the Toolbar.

If that's not the issue, make sure that the Trajectories.dll is there. If it is, please show me a Log File.

Link to comment
Share on other sites

  • 2 weeks later...

New release is up!

Version 2.0.0 for KSP v1.3.1.

Download from GitHub or Spacedock

Full Changelog:

Spoiler
## v2.0.0 for KSP 1.3.1
 - 2018-02-15
### Changes since the last release
 * Completely new and overhauled user interface, based on the DialogGUI system.
   Includes (automated) localizations for Russian, Spanish and Chinese (traditional). If you know any of these languages, please consider helping out with translations. Send all Thanks and Kudos to PiezPiedPy for this amazing contribution!
 * In-Flight trajectory vector line and ground marker persistence bug fixed.
 * ModuleManager updated to v3.0.3
 * When combined with FAR, aerodynamic forces calculated on a packed vessel would lead to NRE's.
   Vessels are now no longer calculated when a vessel is in a packed state. Thanks go to Alex Wang for this bug fix.
 * Higher precision and better performance through more advanced numeric techniques using an RK4 integration method.
   Try disabling the Cache for more precise predictions, predictions that should not kill your FPS anymore.
   Courtesy of fat-lobyte aka Kobymaru.
### For Developers
 * API changes to help catch NRE's from improper calling of the API methods.
 * New API method `HasTarget()` checks if a target has been set.
 * Flickering entries bug in the Profiler is fixed :)
 * Added Reset averages & Show zero calls buttons, a Framerate limiter, avg calls and frame counter to the Profiler.
 
### Known Issues
 * Sometimes In-Flight trajectory ground target marker can cover entire screen when camera is close.

 

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...