2 hours ago, Watsi said:

I actually do wonder, how long MS is planning to provide all these online features for free. Multiplayer, live traffic, live weather, scenery data. Checking my MSFS data consumption: this is a lot of data/traffic (having offline cache deactivated) for quite a lot of users.

I wouldn't be surprised if some of the online services become a payed service at some time in the future.

('Excrement'storm ahead)

I actually do pay already. Monthly, for my gamepass subscription.

4 hours ago, ThomseN_inc said:

Correct me if i am wrong. But wasn't it "the angry mob" who demanded METAR implementation for MSFS? 

Now we got METAR implementation and the angry mob is still unhappy. 

 

Some people really need to get their **** together and stop yelling at the devs for new features and what not every freaking time.

Get a life man.

Welcome to Avsim 🙂

And if you tell it like it is - you get suspended 🙂 That’s how it works. 
 

3 hours ago, MrFuzzy said:

Another thread starting with reasoning and questioning and degrading into insults

YouTubeesque click-bait title? 

6 hours ago, MrFuzzy said:

Watch this video now... the explanation of how complex and elaborate were the weather representation and clouds depiction before SU7 and you can easily realize that what was sold to us in 2020 is no longer there.

It is still there. MeteoBlue weather is as much a part of the sim as it ever was.

But… they are now trying to mix METAR clouds and visibility with the underlying MeteoBlue model weather. Sometimes it works well - sometimes not.

The entire reason they have done this is because of customer demand - because the MeteoBlue forecast sometimes did not match real time METAR cloud reports, and there were substantial numbers of users on the MSFS forums who were asking for airport weather to match the METAR.

Combing two disparate sources of weather data is not easy, and there is much that still needs to be done to refine it further.

I am quite certain that this has absolutely nothing to do with “cost cutting”.

32 minutes ago, JRBarrett said:

The entire reason they have done this is because of customer demand - because the MeteoBlue forecast sometimes did not match real time METAR cloud reports, and there were substantial numbers of users on the MSFS forums who were asking for airport weather to match the METAR.

A reliable metar generated from the Meteoblue forecast model would have been a much better, and I would have thought a much easier to implement, solution than trying to blend the metar with a forecast model. 

1 minute ago, Gilandred said:

A reliable metar generated from the Meteoblue forecast model would have been a much better, and I would have thought a much easier to implement, solution than trying to blend the metar with a forecast model. 

The key word is “reliable”. It is impossible to generate a weather scenario from a forecast (prepared hours in advance) that will exactly correspond to the actual weather that exists at a given place and time. Even weather models that are far higher resolution than MSFS uses cannot do that. That is exactly how Live Weather worked before SU7.

1 minute ago, JRBarrett said:

The key word is “reliable”. It is impossible to generate a weather scenario from a forecast (prepared hours in advance) that will exactly correspond to the actual weather that exists at a given place and time. Even weather models that are far higher resolution than MSFS uses cannot do that. That is exactly how Live Weather worked before SU7.

I flew in South Florida in real life for years. It could be raining on one side of I 95, and sunny two blocks away on the other side. How is the Metar supposed to handle that?

31 minutes ago, Bobsk8 said:

I flew in South Florida in real life for years. It could be raining on one side of I 95, and sunny two blocks away on the other side. How is the Metar supposed to handle that?

METAR is for airports only, and yes, there can be situations like you describe. However there is no other readily-available source of (close to) real-time observations of airport surface weather other than METARS.

Satellite observations and NEXRAD radar could be brought into the mix, though high resolution visible satellite imagery in only available in day time. NEXRAD will show actual precipitation, but is not available everywhere in the world.

1 hour ago, JRBarrett said:

The key word is “reliable”. It is impossible to generate a weather scenario from a forecast (prepared hours in advance) that will exactly correspond to the actual weather that exists at a given place and time. Even weather models that are far higher resolution than MSFS uses cannot do that. That is exactly how Live Weather worked before SU7.

How would it be impossible?  The sim already gets weather data for every location (in cubes) in order for the sim to render clouds, winds, pressure, etc.  Why not just read that as a metar for a given location?

8 minutes ago, Gilandred said:

How would it be impossible?  The sim already gets weather data for every location (in cubes) in order for the sim to render clouds, winds, pressure, etc.  Why not just read that as a metar for a given location?

That is exactly what they were doing until SU7. But that is a prediction of what the might be, based on a forecast. People who use real-world flight planning tools (which use METARS) wanted airport weather to be “exact” for a given airport and time.

 

6 minutes ago, JRBarrett said:

That is exactly what they were doing until SU7. But that is a prediction of what the might be, based on a forecast. People who use real-world flight planning tools (which use METARS) wanted airport weather to be “exact” for a given airport and time.

 

I think you’re misunderstanding me.  The Meteoblue forecast data is what the weather “is” in the sim, not what “it might be,” and I simply want that to be the metar.  I get that it won’t necessarily line up with real world metars.  Would you really be using the current metar for flight planning, or rather a TAF?

6 minutes ago, Rob_Ainscough said:

Weather issues with MSFS are not related to cost cutting. 

I don't visually like the way the clouds are rendered/shaded in MSFS, rarely do they look like real world clouds ... they can certainly be "dramatic" but not realistic.

MSFS weather is "Global" (per video link) which I believe is why they are struggling with weather.  I don't need global simulation of weather (that's just way too much of a processing requirement), I just need simulation of the weather within a Radius of my flight plan or flight path.  As pointed out, there is no way to accurately predict, once you start a global weather instance, where that weather will transform and as such it will not match actual once it's instantiated in the sim over time.

I think what MSFS is struggling with is weather transition from it's initial global state to local reported current state.  Transition is the key and that's what MS/Asobo need to figure out.  I would suggest using a multi-radius implementation such that weather transitions can be smoothed significantly within each defined radius to meet actual conditions without being distracting (i.e. going from heavy clouds to a sudden clear day). 

Per the OPs video link, breaking weather up into squares/cells isn't the way to provide for smooth transitions into accurate weather reports ... and those problems are reflected in the current MSFS implementation.  I hope they do a re-think, nothing worse than going down a path and sticking with it just because that's the path they initially selected.  It's humbling to admit failure and move in a different direction ... but DON'T give up ... nothing wrong with experimentation and getting it wrong, that's how one learns.

I also hope MS/Asobo open up the SDK to allow for adjustments to how the shade/render the clouds ... if users don't like something, open the door for someone else to provide what they do like (be it free or payware) ... open the door leverage the community.

Cheers, Rob.

Weather and Terrain API is number one on the current to-do list. What developers were asking sounds like what you are saying.

https://forums.flightsimulator.com/t/implement-weather-and-terrain-api-s-for-aircraft-developers-to-implement-accurate-radar-predictive-windshear-egpws-and-metar-wind-uplink/442016

44 minutes ago, Gilandred said:

I think you’re misunderstanding me.  The Meteoblue forecast data is what the weather “is” in the sim, not what “it might be,” and I simply want that to be the metar.  I get that it won’t necessarily line up with real world metars.  Would you really be using the current metar for flight planning, or rather a TAF?

Yes, the issue is that a substantial number of people wanted the in-sim weather to be the same as actual real-world weather - not a prediction of what it “might” be.

For flight planning I would use TAF, area forecasts, pireps, airmets and sigmets and any other available information. Operationally, I would use METAR/D-ATIS in flight. I work in a 6 aircraft corporate flight department (in maintenance). All of our aircraft are equipped with ACARS. When one of our aircraft are in flight to a destination the pilots check the latest weather at the destination at regular intervals throughout the flight.

I look at ten weather reports in the morning, they are all different, and by the end of the day half are wrong. 

Once this software gets it figured out better, I'll post a new app, and be able to provide the best service globally!

Here’s hoping they do open up the SDK one day, perhaps then efforts by third parties, for example the 20 years of experience we have combining forecast models with surface observations into a cellular depiction system, could be appreciated and realized in MSFS again.  Maybe having third party weather options (using approved interface methods) is not a bad thing, at least not for actual simmers.  But (enough of) the community is against the idea of any wx addon via SDK, and that will probably have to change before the devs consider it.

Edited January 7, 20224 yr by Damian Clark