1 hour ago, MarcG said:

Because they have no regression testing practices.

I do wonder.. how do you perform regression testing on something like an entire simulator. The datasets must be unimaginable, I personally even know where to start. Can’t even begin to imagine component or E2E.

you would require a sizeable team dedicated to just creating these datasets not to mention the edge cases which imo is impossible to fully cover.

Yes, Asobo’s QC isn’t the greatest but I’d struggle to see most software companies having a full testing suite for something like a flight sim. The test suite for the fbw 320 and 380 is already massive and even then we don’t cover everything and it’s just for the rust side of things. There’s still plenty of in-person required.

just me 2 cents.

 

9 minutes ago, Tuskin38 said:

Not everyone has the issue, so it’s entire possible their QA didn’t either 

It is possible, but for me, it isn't the fact they don't see it or don't have the issue, but what in heavens name they do to cause these regressions, some of which have happened multiple times.  It is slack. 

Some of the bugs are crazy.  Frame rate slider could do with a fix on the next release hopefully (been alright for the last 5 years).  And the curves are still missing when trying to adjust hardware sensitivities etc. 

It does seem that they spend a large proportion of the betas fixing their own regressions.  It really isn't a good look and has become a standing joke in the community and YouTube commentators etc.  It isn't even the Elephant in the room any more.

We all understand betas can be unstable and suffer the odd regression / foul-up that might creep in - it's part of the game, but they really do seem to struggle with a lot of them.

Anyway, I think SU4b has got some way to go yet - I hope, as my SU3 install is doing just great so I don't want an SU4 final that might mess things up.  Better for them to take all the time they want.

I feel like we've gone from one extreme - the FS9, FSX, P3D days of often years without any update and flight sims being on the brink of extinction to the other, where the ground is moving so fast that one has to live in constant fear the next SU or beta build will break the sim. And if that happens you're all out of luck. My journey with MSFS24 so far:

• MSFS24 initial release needs no comment.
• I started to switch from MSFS20 around SU1, as it was getting a bit more stable. It was ok, I accepted it was early days.
• SU2 worked great for me and I didn't even try the SU3 beta. I was a happy camper, enjoying a stable sim.
• Then SU3 was forced on me and completely broke my sim. As I said previously, even a full system reinstall didn't help and there were plenty others with similar issues (stutters, crashes, horrendous FPS, performance degradation over time). I stopped flying altogether for a few weeks as a result.
• Great news! SU4 beta came out, I tried it and it was flight sim heaven. Stable, insanely good performance, all was good for a while. The first 2 updated builds were still good, I was happy even if I could only fly the Fenix and not the PMDG and jetways were missing.
• Build 1.6.11.0 came along and fixed the PMDG - joy! It introduced some random issues with objects floating around airports, jetways floating/sinking and sometimes random texture layers floating at airports. Still, all in all they gave some and they took some and the floating/texture issue was random and usually fixed by a restart.
• Build 1.6.13.0 came along. It fixed nothing I noticed and reintroduced a pre-SU1 issue with corrupted ground textures. This is so notable "in your face" that I'm back to not flying as it's completely immersion breaking and no workaround other than turning off PG (no thanks).

I'm eagerly awaiting the next build that supposedly will fix the texture issue - but what random thing will the new build break in return?

Or maybe the next build will be great and the one after that will reintroduce some game breaking/immersion killing bug again.

And then, with a bit of luck SU4 will be released to the masses, is in good shape and we can all enjoy MSFS for a while. Until SU5 comes along, potentially breaking it all again...and so the cycle starts again. I'm all for progress and I do appreciate the TLC we get with sim updates. But this should be a journey of constant improvements - not one of fear that the next SU is going to break it all again for some people and that fleeting moment of sim happiness is gone again until the next beta/SU. 

I don't know how MS can address this, but in my opinion it needs addressing. Either find a way to quality control SUs/builds so I don't have to live in constant fear what will break next (unlikely) or give us more control over what we want to run. I'd happily go back to SU2 right now, or preferably the previous SU4 beta build whilst I wait for SU4 to mature. But I can't go back to SU3 and SU4 is so immersion breaking for me with the latest build that I'm now grounded for a week (or until the next build comes along, provided it won't break something else...).

3 minutes ago, Speedbird 217 said:

 

I'm eagerly awaiting the next build that supposedly will fix the texture issue - but what random thing will the new build break in return?

 

I know this is said every other day in these threads, but such is the way of the beta. It's not here to be your main day to day way of using the sim. It's here to test a set of very specific fixes to issues.

I do agree that Asobo can communicate this better, and also publish a list of known issues more than they do.

With SU4 a lot of the focus is on WASM, stability and memory management. It wouldn't surprise me if developers  don't really care about how it affects the rest of the sim especially in the early versions of the SUs.

You have an issue or a feature – a new way of managing memory for example. You implement, make sure that part works great, put it out to the beta testers to really make sure it does what it needs to. If it does, great - then make sure nothing is else is broken. That part comes later.

That's why I think Asobo can take a page out of the CIG testing method. Put out notes that say what specifically what is being tested, and what this patch breaks or can be expected to break. Then everyone's expectations are managed properly.

2 minutes ago, Georgleboui said:

I know this is said every other day in these threads, but such is the way of the beta. It's not here to be your main day to day way of using the sim. It's here to test a set of very specific fixes to issues.

I do agree that Asobo can communicate this better, and also publish a list of known issues more than they do.

With SU4 a lot of the focus is on WASM, stability and memory management. It wouldn't surprise me if developers  don't really care about how it affects the rest of the sim especially in the early versions of the SUs.

You have an issue or a feature – a new way of managing memory for example. You implement, make sure that part works great, put it out to the beta testers to really make sure it does what it needs to. If it does, great - then make sure nothing is else is broken. That part comes later.

That's why I think Asobo can take a page out of the CIG testing method. Put out notes that say what specifically what is being tested, and what this patch breaks or can be expected to break. Then everyone's expectations are managed properly.

No and that's fine in principle.

But I'm being forced onto the beta because I am not allowed to go back to SU2, which ran fine, and SU3 is unplayable for me. 

55 minutes ago, Lucky38i said:

I do wonder.. how do you perform regression testing on something like an entire simulator. The datasets must be unimaginable, I personally even know where to start. Can’t even begin to imagine component or E2E.

you would require a sizeable team dedicated to just creating these datasets not to mention the edge cases which imo is impossible to fully cover.

Yes, Asobo’s QC isn’t the greatest but I’d struggle to see most software companies having a full testing suite for something like a flight sim. The test suite for the fbw 320 and 380 is already massive and even then we don’t cover everything and it’s just for the rust side of things. There’s still plenty of in-person required.

just me 2 cents.

 

For me it's the simple things away from the complexity of creating the digital twin that they continually "miss", just look at the number of Mouse related issues for one, the number of blatantly obvious issues that smack your in the face 1 minute into the game etc.

I've never known a developer to struggle so much with mouse bugs, and yet they keep happening, they releasing updates with mouse issues when really someone internally must see it happening.

I totally get it that they can't have a team looking over every single item in this massive game, that's fair enough I've no issues with that, it's the blatant lack of even simple regression testing that they clearly don't possess. Continually every beta something has gone backwards, continually at the end of every beta period not all regressions are fixed, so the bugs list adds up.

There's just sooooo many examples over the past five years and it's extremely tiresome to keep having to find workarounds for issues that should've been caught, many of which are highlighted by simply playing the default game. Many of which would've been caught by a competent regression testing team, which surely with M$ in charge they can at least afford our even sub contract out.

It's frustratingly obvious where they need to improve their back end team, yet five years on its the same old same old and here we are one year on from MSFS2024s release and it's rinse & repeat with the same eye popping in your face issues creeping into betas and public releases, I've seen far far better fun other games development studios.

26 minutes ago, Georgleboui said:

I know this is said every other day in these threads, but such is the way of the beta. It's not here to be your main day to day way of using the sim. It's here to test a set of very specific fixes to issues.

That's where they fall drastically short though, it's practically coding 101 to release a stable beta build to the public testers that doesn't have a bunch of regressions attached too it. The idea of a beta is to fix past bugs, test those scenarios to make sure the fixes have worked, include new features, test those new features for bugs and if any found highlight them to the development team to fix.

In MSFS "flighting" it's a constant swim against the tide as everyone's struggling with new bugs breaking areas that worked perfectly fine for years before, which wastes testers time, it wastes developers time and it creates a slow process in moving forward. 

That's not too say every single regression bug can be found, of course some slip through which is understandable. With this lot though it's the continuous amount of regressions, time and time again that go ignored. It's so obvious I've long since run out of face palm emoticons!

After reading all these comments, i cant help but wonder what the end game is for Asobo? In that, where do they want to the sim to be at the end of all the sum updates? I'm not really sure of the road map to be honest, but how much further will they need to develop MSFS2024 before they call it good? Just curious...

1 hour ago, vonduck said:

After reading all these comments, i cant help but wonder what the end game is for Asobo? In that, where do they want to the sim to be at the end of all the sum updates? I'm not really sure of the road map to be honest, but how much further will they need to develop MSFS2024 before they call it good? Just curious...

In just a poor enough state to be able to sell the next version of MSFS. 

1 hour ago, V1ROTA7E said:

In just a poor enough state to be able to sell the next version of MSFS. 

No

This version is working quiet well on both my MSFS 24 pc’s.

Ground textures look sharp, nice looking clouds and all objects sharp

The only 2 things I encounter:

- when flying with my Prosim 738 is that from 28000 ft and upwards the plane starts to rock up/down. 
The higher I fly the more severe it gets..

- nightlighting does not work correctly on extra external views ( extra view monitor )

On 10/7/2025 at 9:36 PM, Ed_S_Pilot said:

I'm going to give this build a shot. Wish me luck or its back to SU3.

 

Bummer for me, That didn't last long...... Back to SU3, Waiting on the next update.....

 

Blue skies......

WOWW, will this Simulator ever be ready or at least to a state we have for MSFS2020 at now ??

Think i have to wait another year to jump in to the MSFS2024 boat..🤪

cheers 😉

If the vast majority of airport developers continue to release MSFS 2020 versions of new products, I can quite easily see me never switching to MSFS 2024.

There is always a place for a second tier in most things in life.

30 minutes ago, St Mawgan said:

There is always a place for a second tier in most things in life.

Ouch!