METAR does not update wind direction

[MariuszXC]

Having read through a number of topics related to weather in FG I am a bit confused and don't know if what I observed is an expected behaviour or a bug, anyway, here goes the description.

FlightGear version: 2020.4.0
Revision: afe7d996afa7bf1232f43d45fe0280c564680987
Build-Id: none
Build-Type: Dev
SimGear version: 2020.4.0
OSG version: 3.6.5
PLIB version: 185

Environment > Weather > Detailed Weather
Weather conditions: Live data
In advanced settings: Tile selection mode - METAR
METAR received and valid.

FG run by: ./run_fgfs.sh --airport=LGKP --aircraft=c182s

Fligh from LGKP to LGRP

On departure from LGKP property /environment/wind-from-heading-deg contains value consistent with METAR for LGKP (320 deg in this test)

As flight progressed a new METAR from LGRP was received, with the usual 'turbulence' effect. The new METAR was correctly displayed in Phi interface.
Wind direction in METAR for LGRP was 250 deg, which is perfect for RWY 25 there.

However, during approach plane behaved as in a strong crosswind.
Checking the /environment/wind-from-heading-deg revealed, that it still had an old value of 320.

Performing: Environment > Weather > OK (i.e. clicking OK in Weather menu) updated the property with a new value and plane started to behave as expected in a headwind.

So.. is this an expected behaviour or a bug, or does it depend on yet another setting I missed?

[wkitty42]

it is a bug, apparently...

FWIW: someone on the discord is doing work in this area to try to smooth out the winds a little more... no clue if/when they will complete it, though...

[TheEagle]

In case it's relevant, i also had the exact same bug.

[MariuszXC]

Yes, I forgot to mention, but Eagle helped me verify and found the same behaviour on his computer during yesterday's MP event.

[wlbragg]

You should definitely file a bug report so it is on record and has a chance to be fixed.

[MariuszXC]

OK, bug report filed.

[Thorsten]

First question - what was the 'Wind Model' setting?

As flight progressed a new METAR from LGRP was received, with the usual 'turbulence' effect.

Since such a transition effect is generally associated with changing winds, it is hard to understand what would have caused it if the wind did in fact not change - so this is a 'need more info' thing.

Can the issue be reproduced at all? Is there any indication that AW crashed and did not update wind at all?

[TheEagle]

It is reproducible - I had the exact same issue at the exact same point in the exact same flight. We were flying with live METAR on AW. We departed from one airport, and on the way to another a new METAR with a different wind was received (either the old had a crosswind and the new one a head / tail wind or the other way round, I don't remember), but the wind actually blowing in the sim did not change - only after we went to Menu -> Environment -> Weather and clicked Apply, the wind was actually updated in-sim. I was using the constant wind model as for the other wind models AW said that they could not be used with live weather.

[V12]

Some errors in the NASAL console related to AW ?

[TheEagle]

Nope.

[Thorsten]

I was using the constant wind model as for the other wind models AW said that they could not be used with live weather.

That would explain the lack of change, because 'constant' actually means that: take the first wind you get, keep it constant no matter what. Which is what you observed.

The wind model you should have been using is called 'aloft waypoints' and that works with live weather just as well.

(Please take any debug advice from V12 at your own peril - he doesn't understand the internals of AW and hasn't coded any weather code)

[TheEagle]

Thanks Thorsten, I'll try that on the next flight !

[MariuszXC]

First question - what was the 'Wind Model' setting?

Thanks for asking the right question.

If you look at post #1, I wrote that I was not sure if it is a bug or user error:

So.. is this an expected behaviour or a bug, or does it depend on yet another setting I missed?

but given the input from more experienced users I ended up filing a bug report, while it was a case of user error.

Forum posts show that I wasn't the first one confused, and likely won't be the last. Which brings me to a point.
You wrote a complicated piece of simulation, with many options. However there is no weather related help anywhere in sim. Nor is there a Support>Weather section in this forum. My point is a suggestion, that perhaps it would be a good idea to include a help button in the weather related dialog, which would open a browser pointed at our wiki https://wiki.flightgear.org/Weather

This might lessen your frustration with users complaining that something in AW is not working, while all it takes is a pointer to an already existing documentation.

[V12]

(Please take any debug advice from V12 at your own peril - he doesn't understand the internals of AW and hasn't coded any weather code)

I haven't read bigger stupidity than this in a long time.
I assume you have a car with an ESP system. Do you understand the internal processes of this system? Or do You understand shifting strategy of the automatic transmission of Your car ? And in case of transmission with hydrodynamic converter, do You know how works its auto-lock ? When is engaged and when not ? This knowledge is not necessary for you as a user, but you demand that things works perfectly. And when it doesn't works correctly, You can change the car.

[Thorsten]

You wrote a complicated piece of simulation, with many options. However there is no weather related help anywhere in sim.

Well, let's say that the UI design is something that was hotly discussed and then a compromise found - it is not how I would have structured it, and indeed at one point a few years ago I planned to create a modern canvas UI with things like tooltips when I was told that canvas UIs are discouraged...

So yeah, I agree, it's not optimal, but I won't touch it any more, too much red tape for too little gain

As to where people read documentation - difficult topic. It's actually not only on the wiki but also under Documentation/ - at a time it was generally assumed that this is where people would look.

I haven't read bigger stupidity than this in a long time.

Why? Pray tell, this is very amusing.

Or do You understand shifting strategy of the automatic transmission of Your car ?

By and large I do, yes - it's not rocket science (but then, I actually do understand rocket science... )

This knowledge is not necessary for you as a user, but you demand that things works perfectly. And when it doesn't works correctly, You can change the car.

Actually I don't change my car every time something doesn't work perfectly (speaking of big stupidities...) - I take it to the repair shop if something critical isn't working, otherwise I leave it be. It's rather expensive otherwise.

But see - this thread IS the repair shop. And when I have my car there, I shut up and don't tell the car mechanic what to do and how to approach the problem - because I assume he generally knows better than I do - it's his job after all. Should I ever end up having the problem diagnosed by the designer of the car engine - well, he knows FAR better than I do.

But that's not what YOU do - YOU are not the designer of AW, in fact you have a pretty wrong picture of its internals which indicates you never even bothered to try to read the code or the documentation - and yet YOU feel you need to tell people what they should do as if you were competent to diagnose the problem.

You're that guy who tells the car mechanic how to do his job without ever having looked under the hood