Board index FlightGear Development Spaceflight

Space Shuttle - Development

Discussion about development and usage of spacecraft

Re: Space Shuttle - Development

Postby legoboyvdlp » Mon Nov 18, 2019 8:35 am

Ok - well; it worked the second time (after compiling the second bugfix) but I'll make another few launches in that case to be sure.
User avatar
legoboyvdlp
 
Posts: 7981
Joined: Sat Jul 26, 2014 2:28 am
Location: Northern Ireland
Callsign: G-LEGO
Version: next
OS: Windows 10 HP

Re: Space Shuttle - Development

Postby Thorsten » Mon Nov 18, 2019 9:38 am

It's hard to spot - I had two okay launches and then I watched one from external view, when I changed back into the cockpit I found that I exploded, so I pinned that on a weird glitch causing excessive acceleration when the frame gets delayed when changing view - and the instance of the AP failing to engage in TAEM mode I pinned on too high energy - only after eadirt reported similar things I started to connect the dots.

That's the mean thing about this bug - it's scattered instances of something all over the place, and each instance could be something else. But the overall pattern suggests that it is not.

In 2018.3 I actually see zero glitches with the Shuttle - every scenario I've tried worked flawlessly, no oddities at all. So any oddity you observe is potentially a suspect.
Thorsten
 
Posts: 12490
Joined: Mon Nov 02, 2009 9:33 am

Re: Space Shuttle - Development

Postby GinGin » Mon Nov 18, 2019 6:30 pm

Try to revert back to 2018 indeed.
I have played with that one all this year, and I never encountered all the bug that have been mentioned for a few months
GinGin
 
Posts: 1580
Joined: Wed Jul 05, 2017 11:41 am
Location: Paris
Callsign: Gingin

Re: Space Shuttle - Development

Postby legoboyvdlp » Mon Nov 18, 2019 8:26 pm

About the aoa -- this time it didn't exceed limits, but I did notice it was giving the "watch AOA" message after max QBAR at which point it had begun to pitch again, increasing AOA to above 5-6 degrees. At Mach 5, delta inc is 7 degrees; however the guidance manages to decrease this however with increasing x-track.
User avatar
legoboyvdlp
 
Posts: 7981
Joined: Sat Jul 26, 2014 2:28 am
Location: Northern Ireland
Callsign: G-LEGO
Version: next
OS: Windows 10 HP

Re: Space Shuttle - Development

Postby Thorsten » Mon Nov 18, 2019 8:45 pm

I hate to break this to you, but since a while you're responsible for managing launch trajectory - if you have a payload/initial orientation combination which is dangerous, you need to i-load throttle down earlier or longer in the mission file. There's no automated safety check done on the specified trajectory. :D
Thorsten
 
Posts: 12490
Joined: Mon Nov 02, 2009 9:33 am

Re: Space Shuttle - Development

Postby legoboyvdlp » Mon Nov 18, 2019 8:50 pm

Right - thanks :)

I just tried a 2EO abort - it worked rather well, ending up in the Atlantic:
Image

It was pretty close though at 3.7G...

Image
User avatar
legoboyvdlp
 
Posts: 7981
Joined: Sat Jul 26, 2014 2:28 am
Location: Northern Ireland
Callsign: G-LEGO
Version: next
OS: Windows 10 HP

Re: Space Shuttle - Development

Postby eatdirt » Tue Nov 19, 2019 12:29 am

just to say that I gave a try to latest next snapshot with some of the bugfixes implemented, but it looks like some remain. Still got all the
Code: Select all
Nasal runtime error: No such member: new
when I start from an empty .fgfs/ directory, plus another one (maybe I missed it before):

Code: Select all
Nasal runtime error: No such member: condition_manager
  122.28 [ALRT]:nasal        at /home/chris/.fgfs/Aircraft/SpaceShuttle/Nasal/stages.nas, line 3058
eatdirt
 
Posts: 1012
Joined: Wed Aug 15, 2018 3:06 pm

Re: Space Shuttle - Development

Postby wkitty42 » Tue Nov 19, 2019 2:31 am

legoboyvdlp wrote in Sun Nov 17, 2019 8:37 pm:are -4.7 (increasing)

what do you mean by increasing? the numbers are growing toward zero?
"You get more air close to the ground," said Angalo. "I read that in a book. You get lots of air low down, and not much when you go up."
"Why not?" said Gurder.
"Dunno. It's frightened of heights, I guess."
User avatar
wkitty42
 
Posts: 9146
Joined: Fri Feb 20, 2015 4:46 pm
Location: central NC, USA
Callsign: wk42
Version: git next
OS: Kubuntu 20.04

Re: Space Shuttle - Development

Postby Thorsten » Tue Nov 19, 2019 8:15 am

Code: Select all
Nasal runtime error: No such member: new


These are really worrisome - they'd be a sign of Nasal memory corruption of sorts. Basically the instruction finds an object with the correct name of the hash that contains the class in question, but that object is not actually a hash, so it has no instancing member function new().

I have no idea what causes them - algorithmically it's impossible to have the same code running in 2018.3 and not in 2019.2 with this error because the Nasal code is the same. So it has to be a framework problem. I have back then had similar issues with canvas and never been able to track down the cause.

Code: Select all
Nasal runtime error: No such member: condition_manager
  122.28 [ALRT]:nasal        at /home/chris/.fgfs/Aircraft/SpaceShuttle/Nasal/stages.nas, line 3058


In this case not even an object with the name of the hash could be found - usually I get them when there's a parse error such that a file could not be loaded, then the object is not in memory, but I am fairly sure the Nasal of the current snapshot is not afflicted by parse errors.

Basically any of these errors would mean the Shuttle is a no-op.

Sorry, but these do not smell like Shuttle-side issues, that's some deeper framework problem - if it persists, you'd have to report to the mailing list. I'll test the bugfixed 2019.2 today and see what I can observe.
Thorsten
 
Posts: 12490
Joined: Mon Nov 02, 2009 9:33 am

Re: Space Shuttle - Development

Postby Thorsten » Tue Nov 19, 2019 11:08 am

Unfortunately still getting plenty of oddities in 2019.2 - no Nasal errors in the console, but integrator windup in the airfoil roll channel during a TAL (the channel should not be active and hence not integrate before qbar = 10 psf) and - weird, weird - brakes simply failing after touchdown in a second test.

I have no idea how that ties into a pattern, but I've done a week of tests with 2018.3 with zero glitches, so two in as many tests is too much to be comfortable.

Something is going on, and we need to find out what it is.
Thorsten
 
Posts: 12490
Joined: Mon Nov 02, 2009 9:33 am

Re: Space Shuttle - Development

Postby legoboyvdlp » Tue Nov 19, 2019 5:29 pm

wkitty42 wrote in Tue Nov 19, 2019 2:31 am:
legoboyvdlp wrote in Sun Nov 17, 2019 8:37 pm:are -4.7 (increasing)

what do you mean by increasing? the numbers are growing toward zero?


No; they're increasing in magnitude; sorry - I should have made myself clear.

That's incredibly wierd...

The only things that come to mind as significant core changes are:
- subsystem refactoring
- nasal GC changes
- JSBSim changes (seemingly a symptom of worse problems)
User avatar
legoboyvdlp
 
Posts: 7981
Joined: Sat Jul 26, 2014 2:28 am
Location: Northern Ireland
Callsign: G-LEGO
Version: next
OS: Windows 10 HP

Re: Space Shuttle - Development

Postby wkitty42 » Tue Nov 19, 2019 7:18 pm

i want to lean toward JSBsim changes having unseen consequences due to the way we interface with it... i can't see the subsystem refactoring or the GC changes affecting that... well, maybe GC but only if it is cleaning objects that have been released that shouldn't have been... weird stuff, for sure... i know i've seen, over the years i've been here, numerous instances of memory corruption... probably due to buffer overruns from using unsafe copying procedures but that's a WAG at best...
"You get more air close to the ground," said Angalo. "I read that in a book. You get lots of air low down, and not much when you go up."
"Why not?" said Gurder.
"Dunno. It's frightened of heights, I guess."
User avatar
wkitty42
 
Posts: 9146
Joined: Fri Feb 20, 2015 4:46 pm
Location: central NC, USA
Callsign: wk42
Version: git next
OS: Kubuntu 20.04

Re: Space Shuttle - Development

Postby Necolatis » Wed Nov 20, 2019 10:28 am

[quote="Thorsten";p=356841]
Code: Select all
Nasal runtime error: No such member: new


I sometimes get that if I run nasal hashes before fdm-initialized signal has been sent.
If that's the case here, then maybe timing has changed.
"Airplane travel is nature's way of making you look like your passport photo."
— Al Gore
User avatar
Necolatis
 
Posts: 2233
Joined: Mon Oct 29, 2012 1:40 am
Location: EKOD
Callsign: Leto
IRC name: Neco
Version: 2020.3.19
OS: Windows 10

Re: Space Shuttle - Development

Postby Thorsten » Wed Nov 20, 2019 10:41 am

How would FDM-initialized figure? I can understand that it causes potential property I/O issues, but not that it causes Nasal-internal problems.

More strangely, I'm not seeing any of this in 2019.2 myself... So it must be some weird timing issue.
Thorsten
 
Posts: 12490
Joined: Mon Nov 02, 2009 9:33 am

Re: Space Shuttle - Development

Postby WoodSTokk » Wed Nov 20, 2019 11:45 am

IDK how many scripts the Shuttle has, but you can delay the execution of every main-loop with:
Code: Select all
setlistener("/sim/signals/fdm-initialized", func { settimer(kick_loop, 5); });

I think the FDM will not need more than 5 seconds.
WoodSTokk
 
Posts: 1077
Joined: Tue Oct 17, 2017 3:30 pm
Location: Milky Way/Sol/Earth/Europe
Callsign: SX-W57
IRC name: WoodSTokk
Version: 2020.4.0
OS: Debian Bullseye

PreviousNext

Return to Spaceflight

Who is online

Users browsing this forum: No registered users and 4 guests