1 hour ago, bobcat999 said:

You are another one missing the point it seems.  It isn't a fix at all, it is a trial work-around to try to help identify the problem.

That's what you make of it now: the site you linked too clearly offers this as a work-around, period. They aren't trying to help identify problems at all. So I don't think anyone is missing the point of the blog. If "it isn't a fix at all" then why did you call the topic "Performance Fix!"? 😉 It's posted as a work-around and it's a stupid one if you ask me (which you don't, but well... 😉 ). I'd be surprised if this doesn't turn out to be an April Fools' prank. It HAS to be an April Fools's prank!

2 hours ago, tup61 said:

...the site you linked too clearly offers this as a work-around, period. They aren't trying to help identify problems at all. 

Sorry, it was me that posted 'fix' for sure, my mistake.  They actually call it a work-around, and temporary, as they say this.   

'When there’s a hotfix or patch available for the MSFS, you need to return these numeric folders back to their original location'

So sorry once again Tup - I was just trying to help, and nobody had to do it.  Anyway, no doubt other 'April fools jokes' will be around tomorrow.  

14 minutes ago, bobcat999 said:

They actually call it a work-around, and temporary, as they say this.   

'When there’s a hotfix or patch available for the MSFS, you need to return these numeric folders back to their original location'

That's not why they said this: the complete quote is 'When there’s a hotfix or patch available for the MSFS, you need to return these changes explained below to their original states before updating or you might have to download and install the whole MSFS again." So the reason for returning the folders is to prevent an update from going wrong.

But don't get me wrong, I do believe you are seeing this as a temporary work-around and I thank you for the heads up: no problems there and everyone is free to give it a try. But you don't have to defend the blogger for suggesting something rather odd. 😉

OK Sure! 👍  I returned my folders anyway, as it didn't fix the specific Salzburg pausing issue for me.  Asobo are working on more fixes I am sure - Plenty of other places to fly in the meantime.

5 hours ago, MrFuzzy said:

shouldn't affect the framerate when it's closed, but it does. And so does the ATC when the transcript reaches a certain number of lines, even if its window is closed (fortunately there is a community fix for that).

Just wondering if you would know where this ATC fix could be found? Thanks.

2 hours ago, paj said:

Just wondering if you would know where this ATC fix could be found? Thanks.

Here: https://www.dropbox.com/s/ztie75f2cx9vzny/ATCStutteringFix.zip?dl=0

Just decompress the file into the community folder and restart MSFS.

It seems this "tweak" has been removed from flightsimblog pages. Which is probably good. I regret jumping in to soon - I blame the relentless quest for performance (sigh). Even though I moved the default scenery folders back to the original folder (I even took note of the folder location and stored that in a notepad .txt prior to moving the folders), I couldn't get the airports to show in the sim. Hence, I had to resort to a re-install. 

Take home message: Don't try this as it might result in the need to re-install MSFS. Which is not fun. 

@MrFuzzy

Thank you for sharing this link but I find this mod making me having even more questions!!?!

a) Please note I'm not trying to discredit the mod and your contribution, not at all, rather, this simple mod with such a small code change is making me wondering something deeper. Therefore I must ask you first:

Is this mod really helping reducing the load / improving perf on your system?

 

b) I'm wondering something because the only change between the default and this "ATC" mod, is the maximum number of ATC dialog items which are added in a list (from 50 to 4): line 182: var n = data.Dialogs.length - 4;

These items could be visible or hidden (you can scroll the list in this case).

I'm probably not well versed enough in JS, but I can see:

• The code in SetDialogsDatas() is creating a HTML list of LI elements for the last 50 ATC items (4 with the mod).
• The HTML list is recreated from scratch every time this method populates it with items data (it is bound to a listener triggering the call upon an event).
• A list of 50 elements 46 of which are out of view shouldn't take much resources anyhow.
• Even if the caller is not calling  SetDialogsDatas() when 'data' changes, but every frame, this shouldn't be taxing resources much either.

 

c) My initial thoughts:

Presuming the list gets recreated every frame for example, and if this would be known to have an impact on performance:

• I would have kept the list as-is, once constructed, and I'd just change the content, if recreating a new list form scratch is taxing.
• I would have culled out the list to the number of visible items only, if hidden items are still taxing.

Otherwise there is no need for more micro-optimizations like the ones I'm listing, if the list gets recreated only from time to time (only when a new item is added for example).

 

d) My questions:

@cwburnett

You might have much more experience than me with the JS/HTML stack and this is raising some questions to me.

Although JIT JS is nearly as fast (if not faster in some cases) than native x86:

• In your experience, do you find the JS/HTML stack so poorly running, that a list of 50 HTML list items, like the ones on the ATC chat window, is too large and reducing it to 4 could indeed make a difference?
• If this is not the case, do you find CoherentGT is under delivering (compared to others like CEF or WebKit) ?
• Could this otherwise mean using CoherentGT is a bottleneck and is impacting aircraft performance enough to be noticeable?

 

NB: These are genuine questions which this mod in particular is making me wondering from a developer/technology perspective. I'm posting this publicly, instead of just writing you, only because it might help others who are wondering the same kind of questions and are looking for clues. Just to be clear: this is not an attempt to stir the debate about JS/HTML at all and I prefer writing this disclaimer to also make sure there is no misunderstanding.

Edited April 7, 20215 yr by RXP

Great find for me!

Thanks bobcat999! And the blog author:-)

Was hesitant just because I just want to sit and fly(right now) and not tweak more than fly as with FSX. My machine is pretty stout...but
After doing these(except the last 2 items), definitely smoother! Less pauses and stutters!

Thanks!
Cheers,

RB

10 hours ago, RXP said:

@MrFuzzy

Thank you for sharing this link but I find this mod making me having even more questions!!?!

a) Please note I'm not trying to discredit the mod and your contribution, not at all, rather, this simple mod with such a small code change is making me wondering something deeper. Therefore I must ask you first:

Is this mod really helping reducing the load / improving perf on your system?

Absolutely. Without the mod, once the ATC window is populated (let's say after 15-20 minutes of flight, departing from cold & dark at the parking), after every single ATC message I get a 0.5 seconds micropause.

Regardless of settings, current framerate, CPU/GPU load, etc., and with the ATC window open or closed.

The only way to get rid of these micropauses is either disabling the ATC window completely from the upper custom panel or limiting the transcript to 4 entries with the community patch.

This fix has been around for at least 6 months, there may be a better solution but nobody has come out with it for now. 

If you or someone else wants to experiment, you are welcome 🙂

Please also refer to the thread below if you haven't seen it yet:

https://forums.flightsimulator.com/t/atc-related-stutters-appearing-after-flying-for-a-specific-amount-of-time/245442