Great example of Asobo trying to help resolve issues

[polk2 15]

is there a link to a complete download with this fix?

[Steve Dra 7,580]

1 hour ago, DJJose said:

https://www.youtube.com/watch?v=xnVP4cr9X0U

 

By far the funniest movie that ever graced the big screen.  I can't tell you how many times I've watched it....memorized ever line...and like I mentioned in another post when we got to talking about Teri Garr....there is not a bad movie that Marisa Tomei' is in.  She literally stole this show with her lines and her dead-on Brooklyn accent. 🙂

Loved her story about the deer hunting trip when Vinnie asked her about his pants selection!🤣  "BAM"  LOL

[Dominique_K 10,133]

15 minutes ago, Steve Dra said:

By far the funniest movie that ever graced the big screen.  I can't tell you how many times I've watched it....memorized ever line...and like I mentioned in another post when we got to talking about Teri Garr....there is not a bad movie that Marisa Tomei' is in.  She literally stole this show with her lines and her dead-on Brooklyn accent. 🙂

Loved her story about the deer hunting trip when Vinnie asked her about his pants selection!🤣  "BAM"  LOL

 Was thinking about the movie and found that the ticking biological clock thing was actually something I could relate to. Being a ripe 71 yo I hope to see a MFS up to snuff before my clock runs out of time 😅 !. 

Three great actresses, Marisa Tomei in My Cousin Vinny, Madeline Kahn in Blazing Saddle and Young Frankenstein , and "Terry "Put The Candle Back" Garr in Young Frankenstein !   

[MrFuzzy 2,118]

2 hours ago, Dominique_K said:

Over the years I started a couple of threads asking developers if they flew the sim for the own fun. As I recall, there were few answers, generally negative. Only addon developers answered as you can imagine. 

So how can you make sure that something works if you don't use it? Simple, you can't 🙂

Car mechanics drive around the block after they fix a vehicle, even after changing a light bulb you turn the switch on to see if it works before putting away the ladder. 

Instead, for a 120 GB complex simulator accessing 2 PB of Internet data to render the entire planet earth, apparently it's not worth checking the outcome of a 12 GB patch.

We (THEY) need extended beta testing for every new patch, I would be more than willing to help them improve the simulator.

Edited December 1, 2020 by MrFuzzy

[2reds2whites 827]

4 hours ago, psolk said:

If every developer/coder was fired for releasing a code train that broke something in a prior code train there wouldn't be many coders/developers left...  Otherwise no one would need hot fixes would they?  Should LM's developers all be fired for the EA issues?  Come on...

Shoe me any main-studio released software this year that has anywhere near the bugs, issues, and backwards steps that MSFS has. 

You won't find one. Three months in and the developers are actively making some things worse.

We're not talking about a destitute airfield in rural Siberia having an accidental texture problem after a patch. We're talking about shafting an autopilot in a flight simulator.

Edited December 1, 2020 by 2reds2whites

[pmb 5,137]

9 minutes ago, MrFuzzy said:

We (THEY) need extended beta testing for every new patch, I would be more than willing to help them improve the simulator.

Me too. I don't have a problem living at the edge and have quite often been testing stuff before. That's as long as I can return to a stable version if I want to. Which has been working pretty well in X-Plane and AeroflyFS2.

I hope Asobo/MS will realize this early on in 2021 and will keep them nagging on this. (I - and others - did request it already time ago.)

Kind regards, Michael 

[captain420 3,099]

i am grateful for Asobo helping out the community on a fix. But what exactly did they do fix the problem? Does anyone have details on what Asobo and or FBW team did?

[Noel 4,487]

4 hours ago, DJJose said:

Some members get upset when I continue to say please fix the "lighting" issue. 

I give, where do you fly I haven't seen lightening anywhere for over a month of daily flights!  Oh there was before but something changed for me.

[Steve Dra 7,580]

1 hour ago, Dominique_K said:

 Was thinking about the movie and found that the ticking biological clock thing was actually something I could relate to. Being a ripe 71 yo I hope to see a MFS up to snuff before my clock runs out of time 😅 !. 

Three great actresses, Marisa Tomei in My Cousin Vinny, Madeline Kahn in Blazing Saddle and Young Frankenstein , and "Terry "Put The Candle Back" Garr in Young Frankenstein !   

We're like movie-loving twins. 🙂   hehe.....Loved the "put the candle back" scene!   But then again my brain has always been "abby"-something. 🙂 Mel Brooks.....what can you say about the guy....quite the humor-factory!  And as far as he's concerned....."Its good to be the king!" 🤣

Yeah, funny how that "ticking clock" scene becomes more poignant every year at our age in relation to MSFS becoming fully baked before our clock runs out. (I'm just a 59 year-old kid though) 😉  

Edited December 1, 2020 by Steve Dra

[MattNischan 3,778]

21 minutes ago, captain420 said:

i am grateful for Asobo helping out the community on a fix. But what exactly did they do fix the problem? Does anyone have details on what Asobo and or FBW team did?

We've conferred with Asobo also after FBW's report and got details on the issue, which confirmed our suspicions on the cause. Basically, under only a certain set of conditions the lateral nav PID will experience integral windup that will take an extremely long period to overcome, causing asymmetrical bank requests.

The temporary fix is to increase the integral gain to clear the error more quickly, and also lower the value at which it trips windup reset. This is a per plane setting, but they have confirmed a systemic fix will be shipping next update.

-Matt

[marsman2020 1,310]

 

17 minutes ago, MattNischan said:

We've conferred with Asobo also after FBW's report and got details on the issue, which confirmed our suspicions on the cause. Basically, under only a certain set of conditions the lateral nav PID will experience integral windup that will take an extremely long period to overcome, causing asymmetrical bank requests.

The temporary fix is to increase the integral gain to clear the error more quickly, and also lower the value at which it trips windup reset. This is a per plane setting, but they have confirmed a systemic fix will be shipping next update.

-Matt

They shouldn't even be using PID until the last few degrees of the turn.  It should be rolling to achieve a standard turn rate or reach a max bank angle and staying there until it's within ~5-15 degrees of the target heading, and THEN switching to a PID mode only when close to the target heading. I highly doubt any real AP code uses PID for the entire turn, and based on the testing I did with the FSX and XP11 APs they do not appear to do so either.  I suppose you could replicate this behavior by setting the proportional gain such that the commanded bank angle = the max bank angle for any delta in heading > than ~5-15 degrees....

Edited December 1, 2020 by marsman2020

[wthomas33065 994]

Wow.  And all that help and assistance without the need of calling them incompetent.  Amazing.  Shame that some of our own forum members can't display the same level of decorum.  

Just curious.  When was the last time we had such active participation by the development team of any other simulator or previous versions of Microsoft Flight Simulator?

 

[MattNischan 3,778]

27 minutes ago, marsman2020 said:

 

They shouldn't even be using PID until the last few degrees of the turn.  It should be rolling to achieve a standard turn rate or reach a max bank angle and staying there until it's within ~5-15 degrees of the target heading, and THEN switching to a PID mode only when close to the target heading. I highly doubt any real AP code uses PID for the entire turn, and based on the testing I did with the FSX and XP11 APs they do not appear to do so either.  I suppose you could replicate this behavior by setting the proportional gain such that the commanded bank angle = the max bank angle for any delta in heading > than ~5-15 degrees....

The strategy you propose is exactly what they are doing, as of the most recent update, according to Asobo. The AP selects a bank angle based on a physical model that matches the simulator, making the PID only there for small deviations that the formula cannot take into account. However, the integral windup was high enough to swamp the main algorithm.

The frustration is real, and I understand it, but these are very enthusiastic folks and they're working extremely hard. We're all lucky to have such an engaged team, both on the MS side and the Asobo side. Never in the history of Flight Simulator could I, as an outsider dev lackey (no matter how competent or not), just ask the people who are knee deep in the code directly.

-Matt

[Jaxsimmer 147]

Matt,

Great to hear. Of course they are working their butts off in a very strange environment, and its incredibly encouraging they are so responsive to people such as yourself.

I come on these forums every day looking for that pony hidden under a pile of...well you know. It is my humble opinion most of the writers on here really are Larry David under any number of pseudonyms.

Woo hoo! You guys spent $120 bucks on the sim. That is a decent meal out for one with a decent bottle of wine. We all know where that $120 "investment" will be the next morning. For the same amount we have a rough diamond that has some dedicated polishers working on it in a pandemic environment...yet the whining continues.

I can understand frustration with the product, I have it too, but the tone of disrespect for people who are really trying to do their best to give the end user a wow experience gets old after a while.

Bennett

[JimBrown 596]

1 hour ago, captain420 said:

i am grateful for Asobo helping out the community on a fix. But what exactly did they do fix the problem? Does anyone have details on what Asobo and or FBW team did?

I had a quick look at the GitHub pull request for the fix. From what I can see, this is the change that fixed the issue:

In "A32NX/SimObjects/AirPlanes/Asobo_A320_NEO/ai.cfg",

From: headingPID = 2.0, 0.01, 2.0, 0.4, 3.0

To:      headingPID = 2.0, 0.5, 2.0, 0.025, 1.0

Note that this is for the A320 only. Other planes will no doubt require different values.

...jim