Board index FlightGear Release candidates

FG3.7

Release candidate testers are encouraged to post their feedback here. Please read the introduction topic for details.
Forum rules
Please read the introduction topic for details.

Re: FG3.7

Postby Foxtrot15 » Sun Jan 10, 2016 9:37 am

3. Still no particle effects.
not sure about that... i think i saw mention of it on the dev list but don't recall any specific responses regarding it...


So far the newest release (2016.10) still has no particle effects.
Foxtrot15
 
Posts: 379
Joined: Thu Aug 27, 2015 5:01 am
Location: Singapore
Callsign: Ice-99
Version: 2016.1.0
OS: OS X

Re: FG3.7

Postby dtlan201 » Sun Jan 10, 2016 12:14 pm

Yep, Pause button doesn't work
dtlan201
 
Posts: 187
Joined: Sun Sep 29, 2013 3:49 am
Callsign: MIA0774
Version: 3.7
OS: Windows 7

Re: FG3.7

Postby wkitty42 » Sun Jan 10, 2016 6:35 pm

the pause button ("p") works fine over here... at least it did yesterday when i used it numerous times... how is it failing for you??
"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: 5148
Joined: Fri Feb 20, 2015 3:46 pm
Location: central NC, USA
Callsign: wk42
Version: git next
OS: Kubuntu 14.04.5

Re: FG3.7

Postby elgaton » Sun Jan 10, 2016 11:31 pm

I've encountered the same bug. (It's past midnight here, and I need to recompile SimGear + FG, so I'll edit this post in the morning to include some more version information).
  • Start FG with
    Code: Select all
    fgfs --fg-aircraft=<path to the FG aircraft directory> --aircraft=777-200
  • Start the plane via Boeing 777-200 -> Autostart and wait for the procedure to complete.
  • Disengage the parking brake.
  • Increase engine thrust to accelerate along the runway.
  • Press "P" to pause the simulator - note that the plane appears to move forward slowly, even though it should not.

I've tested four planes:

Unfortunately, I don't have the time at the moment to bisect the bug - could you have a look at it, if possible?
NIATCA 2nd admin, regular ATC at LIPX and creator of the LIPX custom scenery
elgaton
 
Posts: 1107
Joined: Tue Mar 19, 2013 4:58 pm
Callsign: I-ELGA/LIPX_TW
Version: Git
OS: Windows + Arch Linux

Re: FG3.7

Postby someguy » Mon Jan 11, 2016 2:10 am

In the new C172P. pressing p-for-pause seems identical to pressing shift-a three times to get 1/8 sim speed. Taxiing or flying, same-same.

Today's build still does this, still lacks particle effects, and still has unresponsive (always white) PAPI/VASI lights.
User avatar
someguy
 
Posts: 1653
Joined: Tue Nov 25, 2008 5:54 am
Location: USA
Version: 2016.2.1
OS: Mac OS X 10.11

Re: FG3.7

Postby wkitty42 » Mon Jan 11, 2016 3:33 am

this really sounds like you guys have the new timing code plus it also sounds like these craft are doing thing in each loop of the frame drawing rather than where they should be being done... i've run into numerous craft that you couldn't leave flying in an accelerated state because their code didn't take the accelerated frame rates into account... looks like someone needs to dig into these crafts' internals and fix these problems since they have now cropped up with the change in the simulator timing mechanisms...
"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: 5148
Joined: Fri Feb 20, 2015 3:46 pm
Location: central NC, USA
Callsign: wk42
Version: git next
OS: Kubuntu 14.04.5

Re: FG3.7

Postby someguy » Mon Jan 11, 2016 6:03 am

wkitty, what models have you NOT seen having pause problems? If there's one I have, I'll try it and report back.

FWIW, I don't see any problems flying at an accelerated sim rate, just the pause failure.
User avatar
someguy
 
Posts: 1653
Joined: Tue Nov 25, 2008 5:54 am
Location: USA
Version: 2016.2.1
OS: Mac OS X 10.11

Re: FG3.7

Postby Thorsten » Mon Jan 11, 2016 7:17 am

i've run into numerous craft that you couldn't leave flying in an accelerated state because their code didn't take the accelerated frame rates into account...


Since that's an unrealistic procedure in the first place, I guess it's up to aircraft developers whether they want to support this or not. In general, it's a lot of numerical work to make guidance algorithms stable under larger timesteps - basically you need to go beyond linear order in the equations, compute higher order correction terms and such like. Messy rather quickly.

I'd frankly prefer if aircraft maintainers focus their energy on improving realism under realistic conditions.
Thorsten
 
Posts: 10434
Joined: Mon Nov 02, 2009 8:33 am

Re: FG3.7

Postby MIG29pilot » Mon Jan 11, 2016 1:21 pm

What happened to time is relevant to the observer? :roll:
User avatar
MIG29pilot
 
Posts: 1454
Joined: Tue May 19, 2015 4:03 pm
Location: 6 feet under Snow
Callsign: MIG29pilot
Version: 3.7nightly
OS: Windows 10

Re: FG3.7

Postby wkitty42 » Mon Jan 11, 2016 8:33 pm

@Thorsten: the problem i spoke of is seen by using accelerated time and having the craft go into up/down wobbling that gets worse and worse... apparently i don't yet understand how the passage of time should affect a craft's movement... one second it is at this XYZ position and the next it is at that XYZ position...

anyway, the fix for the pause problem is in... i didn't see the problem but others have on craft with more/highly detailed FDMs...
"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: 5148
Joined: Fri Feb 20, 2015 3:46 pm
Location: central NC, USA
Callsign: wk42
Version: git next
OS: Kubuntu 14.04.5

Re: FG3.7

Postby Thorsten » Tue Jan 12, 2016 6:02 am

i don't yet understand how the passage of time should affect a craft's movement... one second it is at this XYZ position and the next it is at that XYZ position...


Yeah - and how did it get there? It got there in response to forces acting during that second - but these forces may not have been constant during the second.

Say I'm flying straight and level, and then smoothly pull the stick. The elevator motion is such that at the beginning of the second, the force exerted is zero and at the end it's (for sake of simplicity) full, i.e. elevator-norm is 1.

If you sample the dynamics once per second (Delta t = 1s), the motion of the plane won't show any result of the elevator at all during that first second - because at the beginning of the sampling period the force was zero, so you calculate the timestep with that value.

If you sample with Delta t = 0.5 s, you'll do the first sampling step without force, but for the second step there's at least a force of 0.5 seen, so the plane will respond - albeit in a crude way, hardly smooth.

Which is to say, the longer your sampling step is, the more sluggish the response of the plane to control input and external disturbances will be - and any autopilot not taking this (purely artificial) numerical sluggishness into account will overcorrect and get into an induced oscillation.

Which is to say, if you anticipate to simulate with potentially long sampling timesteps, your guidance algorithms need to be tuned separately to these situations. Or you need correction steps.

The simulation of the Shuttle external tank motion after separation has, if you look at the equation of motion solver, an explicit dependence on the current framerate. There are unphysical correction terms proportional to Delta t which vanish for Delta t -> 0 which compensate for the inaccuracy in the trajectory induced by not sampling the force continuously.

You need to know quite a bit about numerical math to pull these kinds of tricks out of your bag - it's clearly beyond 'just' designing an autopilot logic, which in turn is not something everyone does well.
Thorsten
 
Posts: 10434
Joined: Mon Nov 02, 2009 8:33 am

Previous

Return to Release candidates

Who is online

Users browsing this forum: No registered users and 1 guest