Update Multistage2026 - Development thread

[Matias Saibene]

New release of Multistage2026 for testing.
In this release, I fixed the particle bugs.

 

[gattispilot]

Thanks. What are we testing for? I am looking at the SLS particles test scenario. in 2026 the SLS is missing parts. Not sure why though

[PAYLOAD_1]
MeshName="SLS\deadweight"
off=(0,0,68.695)
Diameter=8
Height=20
MASS=130000
Module="spacecraft\spacecraft3"
Name=Deadweight
SPEED=(0,0,2)
RENDER=1

not sure which .ini needs to be moved to the server?

 

[Matias Saibene]

Thanks. What are we testing for? I am looking at the SLS particles test scenario. in 2026 the SLS is missing parts. Not sure why though

[PAYLOAD_1]
MeshName="SLS\deadweight"
off=(0,0,68.695)
Diameter=8
Height=20
MASS=130000
Module="spacecraft\spacecraft3"
Name=Deadweight
SPEED=(0,0,2)
RENDER=1

not sure which .ini needs to be moved to the server?

In my installation, the SLS appears complete. Perhaps I've missed a file.

Could you please post the Orbiter.log file?

 

[gattispilot]

Here you go. It shows lots of missing meshes but they are there? in the 2026 the pad is not seen. I modified the SLS particles scn

 

[Matias Saibene]

Here you go. It shows lots of missing meshes but they are there? in the 2026 the pad is not seen. I modified the SLS particles scn

Oops! The pad disappeared again!
The pad should be generated by the Multistage code itself according to the scenario options.
I apologize for the inconvenience. I'll be back soon and hope to fix these issues.

 

[Matias Saibene]

Testing Multistage2026 on Linux. I think Linux is stricter than Windows, and memory errors quickly become apparent.
I was able to get the SLS exhaust texture to display. And I also got @Buck Rogers' N1 working (by the way, what a magnificent 3D model!).

 

[Matias Saibene]

I removed several arrays of 7500 elements each and replaced them with std::vector. I'm still having nightmares...

So the question arises: how did MS2015 work perfectly in first place, and then, due to a few minor changes, it no longer works?*
Or better yet, how did it work perfectly with so much uninitialized memory, dangling pointers, and structures the size of an aircraft carrier?

I want to clarify that I'm in no way criticizing the previous work; MS2015 is an extremely complex work of art, but I'm honestly amazed by the mysterious ways of programming in C++.

May God, the Probe, and Bjarne Stroustrup have mercy on my soul!

*It doesn't work on Windows; on Linux, I'm managing to run several rockets (SLS and N1).

 

[Matias Saibene]

Multistage2026 source code for Linux uploaded!
https://github.com/MatiasSaibene/Multistage2026_for_OpenOrbiter/tree/master/Multistage_for_Linux

 

[N_Molson]

I removed several arrays of 7500 elements each and replaced them with std::vector. I'm still having nightmares...

So the question arises: how did MS2015 work perfectly in first place, and then, due to a few minor changes, it no longer works?*
Or better yet, how did it work perfectly with so much uninitialized memory, dangling pointers, and structures the size of an aircraft carrier?

I want to clarify that I'm in no way criticizing the previous work; MS2015 is an extremely complex work of art, but I'm honestly amazed by the mysterious ways of programming in C++.

May God, the Probe, and Bjarne Stroustrup have mercy on my soul!

*It doesn't work on Windows; on Linux, I'm managing to run several rockets (SLS and N1).

Remember that the CPU has to parse all that code every frame, so it is important to optimize things.

 

[Matias Saibene]

After getting frustrated with not being able to make any progress with MS_Pad, I decided to focus on MFD.
The result? MFD isn't displaying, but MS_Pad did appear, and it seems the other errors were mysteriously/magically/miraculously fixed.
And since I used those three magic words to describe what I don't understand, I think I deserve the WWP award:

World's Worst Programmer.

Joking aside, I plan to port the Multistage2026 code from Linux to Windows and soon (500 years later) we'll have dll libraries so you can help me test it.

 

[gattispilot]

Let me know when you got a test version

 

[Matias Saibene]

Let me know when you got a test version

Thanks again!

I'm currently modifying the MFD to use clbkGeneric as the method to communicate with the Multistage2026 rocket. Previously, a pointer to the class was used, but because the class wasn't loaded into memory, the MFD was failing when starting Orbiter and before starting Multistage2026.

 

[dennis.krenz]

Thanks again!

I'm currently modifying the MFD to use clbkGeneric as the method to communicate with the Multistage2026 rocket. Previously, a pointer to the class was used, but because the class wasn't loaded into memory, the MFD was failing when starting Orbiter and before starting Multistage2026.

Yes, I think that is the better way to handle the Orbiter vessel lifecycle. Also, this makes it easier for you to control which interface the vessel provides to this MFD or another MFD, allowing a looser coupling between the releases.

 

[Matias Saibene]

I have another test release.

This time I fixed MS_Pad, which was not displaying due to memory corruption.
What I did was port the code to Linux, and once there, using tools like Valgrind and GCC ASan, I was able to correct many pointers and out-of-index accesses.

However, problems persist:

1. Exhausts are not displaying in all stages.
2. The autopilot is still not working.
3. Multistage2026_MFD is still under repair.

Please let me know if there are any other issues in this release.

 

[Buck Rogers]

Quick test with the SLS.
No CTD!
Launchpad exhaust working. HUD display working.
Booster engine effects/particles not showing properly.
Booster sep good but the interstage seperated at the same time.
No rot drag (RCS)?
No 2nd stage ignition or fairing sep. (J)Jettison not functional.
Otherwise looks very smooth, no lags, jitters etc...

 

[Matias Saibene]

Quick test with the SLS.
No CTD!
Launchpad exhaust working. HUD display working.
Booster engine effects/particles not showing properly.
Booster sep good but the interstage seperated at the same time.
No rot drag (RCS)?
No 2nd stage ignition or fairing sep. (J)Jettison not functional.
Otherwise looks very smooth, no lags, jitters etc...

Thank you so much for your report!
I'm glad to hear that something is working!

I'm going to take a short break from programming Multistage2026 while I read your reports and focus on what the next steps will be.

 

[gattispilot]

Looks good for me. Crawler worked. Not sure how the rocket gets attached to the pad though?

I tried mY sls. I get this in the HUD/mfd

Do the boosters work? It lifted up but no exhaust from them

 

[Matias Saibene]

Looks good for me. Crawler worked. Not sure how the rocket gets attached to the pad though?

I tried mY sls. I get this in the HUD/mfd

Do the boosters work? It lifted up but no exhaust from them

Thanks for testing!

Yes, the exhausts issue is still a bit mysterious; I don't understand why it sometimes works and sometimes doesn't. I have all the variables initialized.
The boosters are working now, but the MFD still isn't. I recommend disabling it because it causes memory corruption.

 

[Buck Rogers]

Yes, the exhausts issue is still a bit mysterious; I don't understand why it sometimes works and sometimes doesn't.

I got the impression that the booster particle effects were running on default (minimal/no settings)- very small and no expansion?

 

[gattispilot]

Trying to get the default SLS to launch. I noticed this in the save scn. notice the save stage and alt steps,....
SLS:Multistage2026
STATUS Landed Earth
POS -80.6538895 28.5898809
HEADING 336.00
ALT 57.434
AROT 151.087 -8.198 28.503
ATTACHED 0:0,MS_Hangar
PRPLEVEL 0:1.000000 1:1.000000 2:1.000000
NAVFREQ 0 0 0 0
XPDR 0
CONFIG_FILE Config\Multistage2026\SLS.ini
GUIDANCE_FILE Config\Multistage2026\Guidance\SLS_57302.65_GNC.txt
CONFIGURATION
COMPLEX
HANGAR
CURRENT_BOOSTER 1
CURRENT_STAGE 1
CURRENT_INTERSTAGE 1
CURRENT_PAYLOAD 1
FAIRING
MET -30.000
STAGE_IGNITION_TIME 0.000000
STAGE_STATE
TELEMETRY_FILE Config\Multistage2026\Telemetry\SLS_57310.61_TLM.txt
ALT_STEPS %.1f,%.1f,%.1f,%.1f
PEG_PITCH_LIMIT 35.000
PEG_MC_INTERVAL 0.100
END