Well, there are people that know quite a bit in theory, but actually coding it to model the theory is harder than just writing down the math. So it's a duality in synchronicity of trying to translate something in multiple dimensional overload to where your brain might disconnect from reality temporarily and say "that seemed right", but it just never got quite to be as good as it could have been. Need the savant, must find master (wow it's late, what am I talking about, think this thread turned my brain to mush).

1 hour ago, eslader said:

They're trying to build a next-gen sim when the previous sim that they have to reference would have been 5 generations old if development had continued. In other words, they're trying to build an SR-71 by extrapolating from the Wright Flyer. They can get there, but it's not going to happen overnight.

In one of the latest videos, an Asobo developer actually said that just because some of them had now taken flying lessons did not mean they were suddenly experts on all the aircraft of the world, and that they were going to need help. At the time, I thought it was a fact based plea to the community for some patience and understanding.

That apparently many of us are not inclined to give them,

Edited October 17, 20205 yr by HiFlyer

Just now, HiFlyer said:

In one of the latest videos, an Asobo developer actually said that just because some of them had now taken flying lessons did not mean they were suddenly experts on all the aircraft of the world, and that they were going to need help. At the time, I thought is a fact based plea to the community for some patience and understanding.

that right there is realllllyyyy hard for some to understand . 

They would have contracted one of the experts if they wanted to do so, you cannot have random members of a community involved that deeply in something so complicated without some type of formal commitment or payment structure, unless they wanted to make it open source I suppose. I haven't done any aircraft or tried any physics, so no idea on what all it will take for the third-party devs to fix it, guessing some magic, no idea.

 

Edited October 17, 20205 yr by SceneryFX

10 minutes ago, rtodepart said:

that right there is realllllyyyy hard for some to understand . 

I paid for a product, I expect the product to work.  I don't owe Microsobo any sympathy if they chose to deliver a half-baked product that doesn't work.  If they needed to hire an actual test pilot to give them feedback on their flight models, then they should have done that, not appealed to the masses to somehow vote on what the flight model should be.  They even had that fun partnership video with Textron right after launch - I wonder that Textron's test pilots think of how their aircraft are represented in the sim?

13 minutes ago, SceneryFX said:

Well, there are people that know quite a bit in theory, but actually coding it to model the theory is harder than just writing down the math. So it's a duality in synchronicity of trying to translate something in multiple dimensional overload to where your brain might disconnect from reality temporarily and say "that seemed right", but it just never got quite to be as good as it could have been.

The 6DOF equations of motion of a rigid body are actually not that complicated. 

On top of that you need aerodynamic code to calculate what the forces and moments are to be summed up, given the current atmospheric conditions and the geometry of the aircraft.

From that the static and dynamic stability coefficients are a logical derivation from the rest of the math.

There is 0 excuse for shipping something with "flight simulator" in the name where you can't even trim the aircraft for straight and level flight, which is the most basic of all flight operations. 

They should have had dynamic stability working from day 1 with a generic aircraft model long before they had weather, or detailed 3D models with rivet-level details, or navigation data, or anything else.

I can see where for some people it might seem like some daunting and impossible task to build a working flight model that we need to give them all kinds of space and time to finish, but the reality is most of what is needed is in like 4 textbooks that any one of hundreds of engineering students is capable of reading and understanding. 

The fundamental problem is that this software was developed as a "world simulator" first with warmed-over FSX flight models bolted onto it.  Vs having the flight models developed first, and then adding in the visual environment later.

Edited October 17, 20205 yr by marsman2020

12 minutes ago, rtodepart said:

that right there is realllllyyyy hard for some to understand . 

Here is the actual quote. At time 36.33

 

8 minutes ago, marsman2020 said:

I paid for a product, I expect the product to work.  I don't owe Microsobo any sympathy if they chose to deliver a half-baked product that doesn't work.  If they needed to hire an actual test pilot to give them feedback on their flight models, then they should have done that, not appealed to the masses to somehow vote on what the flight model should be.  They even had that fun partnership video with Textron right after launch - I wonder that Textron's test pilots think of how their aircraft are represented in the sim?

I'm going to have to assume you did not watch the developer videos, or that if you did, you did not pay attention when they spoke of their internal testing team and some of the particulars.

Edited October 17, 20205 yr by HiFlyer

4 minutes ago, HiFlyer said:

I'm going to have to assume you did not watch the developer videos, or that if you did, you did not pay attention when they spoke of their testing testing team and some of the particulars.

I watched it live and regardless of what they said, I still don't think I owe them any sympathy for delivering a non-working product or any of my own time spent fixing their non-working product.

In fact the fact that they have 40 full time QA people gives me even LESS sympathy for them.  I am finding bugs left and right as 1 person who has spent <100 hours in the sim.  With 40 people working full time that is 1600 hours per week, they should be able to identify and triage many of these bugs with that much testing time.   The only conclusion I can draw is that they are not able to use those QA resources effectively for some reason (poor prioritization, unclear test objectives, or no clear design intent/documentation on how things are actually intended to work for the testers to write bugs against).

Professional test pilots are a thing and I'm sure there are many who would be more than happy to test MSFS flight models in return for adequate and appropriate compensation.  Since all of the aircraft are under license agreements with the aircraft manufacturers, I am sure Microsobo could ask the manufacturers for recommendations on who they could pay to test each aircraft.

Edited October 17, 20205 yr by marsman2020

3 minutes ago, marsman2020 said:

I watched it live and regardless of what they said, I still don't think I owe them any sympathy for delivering a non-working product or any of my own time spent fixing their non-working product.

It seems that you are going to have a long, unhappy wait, then.

10 minutes ago, marsman2020 said:

I wonder that Textron's test pilots think of how their aircraft are represented in the sim?

when that video was posted here i replied and said what im going to repeat now . My few interactions in real life with the guys at textron had been nothing but really  understanding professional interchanges I personally dont think they just made a video to smile at the camera putting the textron  name out there with no intention to have at some point a good representation of their product in FS2020 . but again is a thing here at avsim for some to sometimes type something and assume they know something even after developers themself had clarified their intentions  are the opposite ... 

Just now, rtodepart said:

when that video was posted here i replied and said what im going to repeat now . My few interactions in real life with the guys at textron had been nothing but really  understanding professional interchanges I personally dont think they just made a video to smile at the camera putting the textron  name out there with no intention to have at some point a good representation of their product in FS2020 . but again is a thing here at avsim for some to sometimes type something and assume they know something even after developers themself had clarified their intentions  are the opposite ... 

Microsoft had a lot of intentions that they spoke of with Microsoft Flight! as well.  Intentions =/= results. 

39 minutes ago, marsman2020 said:

I paid for a product, I expect the product to work.  I don't owe Microsobo any sympathy if they chose to deliver a half-baked product that doesn't work.  If they needed to hire an actual test pilot to give them feedback on their flight models, then they should have done that, not appealed to the masses to somehow vote on what the flight model should be.  They even had that fun partnership video with Textron right after launch - I wonder that Textron's test pilots think of how their aircraft are represented in the sim?

The 6DOF equations of motion of a rigid body are actually not that complicated. 

On top of that you need aerodynamic code to calculate what the forces and moments are to be summed up, given the current atmospheric conditions and the geometry of the aircraft.

From that the static and dynamic stability coefficients are a logical derivation from the rest of the math.

There is 0 excuse for shipping something with "flight simulator" in the name where you can't even trim the aircraft for straight and level flight, which is the most basic of all flight operations. 

They should have had dynamic stability working from day 1 with a generic aircraft model long before they had weather, or detailed 3D models with rivet-level details, or navigation data, or anything else.

I can see where for some people it might seem like some daunting and impossible task to build a working flight model that we need to give them all kinds of space and time to finish, but the reality is most of what is needed is in like 4 textbooks that any one of hundreds of engineering students is capable of reading and understanding. 

The fundamental problem is that this software was developed as a "world simulator" first with warmed-over FSX flight models bolted onto it.  Vs having the flight models developed first, and then adding in the visual environment later.

Lol,  you think just dropping some formulas from an aerodynamics textbook is what you do to create a flight model...  welcome to the 1982 MSFS version 1 flight modeling.  Seriously get out of here with that.  Theory and formulas in textbooks is great for a basic understanding of physics, but ignores or trivializes all the real world variables that you have to account for if you are going to simulate them.  (probably a main reason why Xplane has such a basic weather system)  computational fluid dynamics is really really really not simple to get close to real world performance, and doing even a fraction of that in a consumer level product, on an average end user machine is not simple at all.

6 minutes ago, ShawnG said:

Lol,  you think just dropping some formulas from an aerodynamics textbook is what you do to create a flight model...  welcome to the 1982 MSFS version 1 flight modeling.  Seriously get out of here with that.  Theory and formulas in textbooks is great for a basic understanding of physics, but ignores or trivializes all the real world variables that you have to account for if you are going to simulate them.  (probably a main reason why Xplane has such a basic weather system)  computational fluid dynamics is really really really not simple to get close to real world performance, and doing even a fraction of that in a consumer level product, on an average end user machine is not simple at all.

Physics is physics.  I'm sure that there's a bunch of equations from textbooks in the sim right now.  I also think they have used fixed fudge factor coefficients in areas where their math was off which causes things like broken dynamic pitch instability.

Edited October 17, 20205 yr by marsman2020

8 minutes ago, marsman2020 said:

Physics is physics.  I'm sure that there's a bunch of equations from textbooks in the sim right now.  I also think they have used fixed fudge factor coefficients in areas where their math was off which causes things like broken dynamic pitch instability.

Yes and no, here is the issue when coding physics, it's isolation in testing the independent variables without making it so complicated that you end up with a list of hundreds of test cases, as you cannot truly test the results of the modeling until the full physics are all working together simultaneously, because it all interacts. It is something I seriously wouldn't even attempt, been coding for a long time, and I have some background in stats, wouldn't even try to work on it, need a real specialist usually.

In a textbook, it is automatically in an isolated equation only showing you one physical interaction, the problem is trying to cross-plug them all together in a way they all interact simultaneously. Definitely not for the faint of heart. A lot of stuff is not in a textbook I imagine, as some of it might be only proprietary knowledge because every plane has different physics. 

So the first problem was hard enough to solve, but now you have different planes with different physics, that's definitely not a fun issue.

Edited October 17, 20205 yr by SceneryFX

Just now, marsman2020 said:

Physics is physics.

right. and what happens in mathematics or physics when you increase the amount of variables?