Just to give my point of view. FS2020 has great graphics, but it is not really thaaat much difficult implementing good looking objects to a game. After all, in UE4, you just have to click import and tweak the materials to get the same results.

 

I do programming in my spare time, I have built for fun a basic ping pong game that hits the wall and shares the same style with the first video game. So, I wanted to implement an AI for fun. I am right now at 950 lines of code, very very much for such a simple thing. Still the game is not working 100%. The AI is supposed to be unbeatable, but, if you try the game, you will see you can sometimes loose.

 

I have corrected a lot of mistakes, but still there are mathematical errors and they are simple just my faults. Programming is so much complex, Asobo is ruining new things with every update, because I think they are using the legacy code that was built by people with much more skills then they use to have.

 

If you want to try my primitive game, it`s there (development build, so not good looking):

https://drive.google.com/file/d/1Nc8eLMobVLYk80X9CPNS8kU8h_C6qbGz/view?usp=sharing

Can we all just step back and take a breath? 

1.  It's obvious that the A/P behavior is most likely a mathematical error that was introduced when they tried to correct some of the oscillation behavior that people were complaining about with the A/P.   It seems to affect all aircraft, and really deadens the A/P.  I have no doubt it will be corrected.  

2.  This is NOT some intentional "dumbing down" of the flight model, because quite frankly, that makes no sense.  Yeah, Asobo decided to make the A/P LESS reliable and LESS able to track the Localizer so that the Xbox gamer can play the game easier.  Sorry, don't buy it.

3.  About "dumbing down".  There's a lot of debate about that term, and I think the more accurate term is to simplify some of the instrumentation and avionics.  For example.  In the default A320Neo, you can click on the destination and choose your approach at any time in the flight plan, where as, I hear that in the real one, you can't do so until you either manually enable approach mode, or the plane passes the DECEL checkpoint on the flight plan.  Or not requiring the player to wait 7 minutes to align the INS prior to continuing the flight.  It's those design choices that I have no problem.  Not something silly like allowing a 747 to fly inverted under the Golden Gate bridge.  But certain areas of the flight model COULD be made easier for players, such as more relaxed "overspeed" behavior, or relaxed stall behavior which could be set by a realism setting.

That being said, we are all aware of the frustration that we feel when updates appear to break things that were not broken, and that is understandable.  But I also know that none of us know one thing about the actual dynamic at Asobo or Microsoft at this time.  We have plenty of armchair QB's and Developers ready to give their "expert opinion" on software development who also know squat about what is really going inside both software studios at this time.  The fact that they even released product during the pandemic is pretty impressive to me.  

It's safe to assume that access to testing systems has been reduced substantially.  This isn't the type of game you can adequately test on a VM.  No, this has to be set up on multiple rigs, and that means getting people into office spaces.  We also don't know how many people are assigned to QA or how their release schedule is being driven.   But one thing I can tell you is that COVID-19 has disrupted EVERYONE's life for nearly 9 months now, and it is silly to think that this hasn't impacted this particular release.

So while we can all be frustrated, I suppose we can also give the development team a bit of a break.  After all, it is, in fact, only a game.  

 

Game or not there is a thing called Quality Assurance.

https://forums.flightsimulator.com/t/autopilot-behavior-not-linear-using-hdg-or-nav/326978

 

That's the thread to vote on the AP issue.

10 hours ago, Steve Dra said:

So the possible workaround with the rolling cache did not appear to work for me. 

Even a broken clock is right twice a day, and if people aren't doing controlled testing with the same flight plan, speeds, and procedures, they might get different results, and then 'blame it on the rolling cache".

 

22 minutes ago, wthomas33065 said:

I have no doubt it will be corrected.

I am sure too, but when? Are we going to get this fix, soon or will we need to wait a month for it?

28 minutes ago, Adrian123 said:

Game or not there is a thing called Quality Assurance.

 

6 minutes ago, bendead said:

I am sure too, but when? Are we going to get this fix, soon or will we need to wait a month for it?

Who knows.  But this is a year where I've had to wait over a month to buy toilet paper or paper towels, so I think I can bear to wait a month for a fix for a piece of entertainment software.

It's called perspective.   

Edited November 29, 20205 yr by wthomas33065

I don't think it's because of being remote, I'm sure they all have mega PC's at their homes, most of them anyhow. So its just a matter of Github'n a new pull request and pushing the new code. I would surmise that working from home has caused a lost in productivity though, because people get distracted. If they are working from home, who knows.

 

3 minutes ago, wthomas33065 said:

It's called perspective. 

It’s called YOUR perspective. I have a slightly different perspective, as do so many others it seems. Your lack of toilet paper does justify the backwards steps this sim keeps taking. 

Just now, RaptyrOne said:

It’s called YOUR perspective. I have a slightly different perspective, as do so many others it seems. Your lack of toilet paper does justify the backwards steps this sim keeps taking. 

And your issues with the autopilot on your toy airplane game doesn't justify putting people at risk as infection levels continue to climb worldwide.  Sorry about your "inconvenience".  (smh)

3 minutes ago, wthomas33065 said:

And your issues with the autopilot on your toy airplane game doesn't justify putting people at risk as infection levels continue to climb worldwide.

You are aware, that in the world we live in today, we can work at home, something that was used way before covid for software development.

11 minutes ago, bendead said:

You are aware, that in the world we live in today, we can work at home, something that was used way before covid for software development.

I am also aware that QA'ing entertainment software is NOT something that can be easily done at home.  I spent time as a games tester under contract at Microsoft for Close Combat III The Russian Front.   The labs we had setup and the number of configurations we had to cycle through were mind boggling.  UI testing alone took days behind one way glass with staff taking notes asking people to navigate the UI and see where they mess up.  Endless hours comparing artwork with the underlying map data to make sure each building had the right amount of "windows" and that ridge lines and other topological data matched the artwork.  Many sleepless nights stress testing the product, both in single player and multiplayer.  And CC3 was no where near the scope of this product.

And some of the hardest work was post release because, most of the testing staff are sent to other projects under development, and a skeleton crew of testers are left to QA "continuing development".  So now those folks have to ramp up on areas of the game that during development they never focused on.  

The fact is, we DON'T know what the situation is at Asobo or Microsoft, but we comment and criticize as if we do.  I'm simply saying, I don't know the answers, so I'm willing to give them the benefit of the doubt.  

Edited November 29, 20205 yr by wthomas33065

These are not those types of "Hidden" issues, these are blatant bugs. There are millions of game devs working from home right now, most are.

I still don't understand why they changed the night lighting, it was ALMOST perfect (Except for the orbs) in the original release. The autopilot is probably someone pushed the wrong code or someone changed something and didn't realize the bugs.

They simply are not testing this at all. Any tester could have found these bugs from a basic take-off, landing test, I mean this is a Flight Sim, if you don't check the landings with the AP, what are you checking...

 

6 minutes ago, wthomas33065 said:

I spent time as a games tester under contract at Microsoft for Close Combat III The Russian Front.

That was 22 years ago, a lot changed since.

6 minutes ago, wthomas33065 said:

so I'm willing to give them the benefit of the doubt.  

We did, first 3-4 patches with problems, it was not looking great, but we gave them time to improve.

Right now, after 3 month, we still have bugs, that could have been detected just by doing a 30min flight, and there is no improvement on that after all these patches.

12 minutes ago, SceneryFX said:

These are not those types of "Hidden" issues, these are blatant bugs. There are millions of game devs working from home right now, most are.

I still don't understand why they changed the night lighting, it was ALMOST perfect (Except for the orbs) in the original release. The autopilot is probably someone pushed the wrong code or someone changed something and didn't realize the bugs.

They simply are not testing this at all. Any tester could have found these bugs from a basic take-off, landing test, I mean this is a Flight Sim, if you don't check the landings with the AP, what are you checking...

 

First of all, you are assuming that the bug is 100% repeatable, all the time.  I have found this not to be the case.  Since the update alone, I have had issues where the AP in the A320 works almost normally, save for the fact that FL changes between FL200 and FL300 seem painfully slow.  The plane follows the flight path, it follows the approach path and captures the localizer.  Brilliant. 

Then the next flight, it ignores the flight path completely until I enter the next waypoint as a Direct To setting.

Then after that, I take off, enable the autopilot and the autopilot puts me into a steep bank, overspeeds the aircraft and almost crashes me to the ground if I don't quickly engage the selected altitude mode on the AP.

And that's ONE plane.  Three different behaviors, Three different flights.  I've been testing this for the last 3 days on the A320 alone, and I can't get 100% reproducible steps.  In most cases I can see there's "a problem" but the behavior is different each time.  

The A320 problems also differ when you start C&D vs when you start at the runway.  Yet, who would think that would play into it?  So if they test starting at the runway, and then miss the C&D test....you can start to see how the variables stack on each other.

So the idea that you say they are simply not testing at all, comes from a lot of assumptions.  The "bug fix" was to fix the oscillations of the autopilot.  So the most direct test would be to start in flight, and engage the autopilot.  Do the autopilot oscillations stop?  Yes.  Test passed.  Next test.  

How long does it take to do a real C&D test for takeoff and landing?  45 minutes?   an hour?  times how many aircraft?  On a bug, that on the same aircraft can show 3 distinct different behaviors?

So I can understand how it can be missed, and I can understand how it can be tough to find.

I recently reported a conflict with two scenery products that produce a 100% reproducible crash to desktop on my machine when those are the only two products in the community folder, and the CTD disappears when one of them is removed.   Both developers have stated that they can't reproduce, and apparently my system is weird because no one else has confirmed it on their system.  So I get how hard it is even when a person can report a 100% reproducible behavior on their system.  

Edited November 29, 20205 yr by wthomas33065