Clearing app cahe

[iHolophyte]

Cheers skull

So to recap. When I reboot it clears what ever it is it needs to clear. Lol

We'll leave it there for me. I'm not a massive techie.
Interesting topic though.


iHolophyte

 

[pianoman]

LOL. Ok.

Lets list off the most common ones then for a *nix based OS. I believe 95% plus of what I am about to type applies to both Android and iOS for reference.

1) CPU has a cache.

2) The OS has several caches. This is due to almost everything being kept as a file on disk.
a) programs in memory
b) drivers in memory
c) data for flash memory access

3) Apps has a cache launching area. Android has this for sure because of the Dalvik Virtual Machine. I don't believe iOS has one because all apps are native. PC have one for optimizing launches in the future.

4) Each App has a data section in flash memory available to it. Calling it a "cache" is sort of a misnomer though. It would be better called a "scratch pad" and "permanent storage".

So the following are cleared on reboot - 1 and 2. 4 is dependent on the app/developer. 3 is handled when you load an updated version of the app.

Thanks, Skull! I knew you'd eventually chime in and bring order from chaos. You didn't come right out and say it in your recent posts, so let me be sure that you "do" agree that the app switcher has absolutely nothing to do with creating availability in RAM. Agreed?

I think Apple designed the switcher as a mini shortcuts window to easily get to your most used apps and I like it as a design feature.

I think it was you yourself, Skull, that termed the belief, that the app switcher could release ram, as a "myth".

Thanks!

 

[Skull One]

Thanks, Skull! I knew you'd eventually chime in and bring order from chaos. You didn't come right out and say it in your recent posts, so let me be sure that you "do" agree that the app switcher has absolutely nothing to do with creating availability in RAM. Agreed?

I think Apple designed the switcher as a mini shortcuts window to easily get to your most used apps and I like it as a design feature.

I think it was you yourself, Skull, that termed the belief, that the app switcher could release ram, as a "myth".

Thanks!

Wow. This is where one of my posts can be taken EITHER way for the argument at hand. And that is because of "how things work" in iOS.

Technically what I said was that the Task Bar can be used to kill an app. IE force it to reload from scratch. And that the task bar survives a reboot because it is technically a list of last used apps. BUT...

What it leaves in memory when you do that, that is entirely different subject and one I should have been a lot more evidential about in my last major discussion on the subject. For that I apologize.

If you read my post #42 in reference to the pictures I posted on #39 you will see beyond a shadow of a doubt that the task bar can be used to free memory (IE I ain't jailbroke, so I can only kill an app that way).

BUT the issue is "which apps" when you kill them show this behavior. IE the Mail or Message apps won't show this. Tiger Woods 2012 on the other hand does in fact show what iOS does when the task is removed from the Task Bar. The reason for the difference is the amount of memory Mail takes up is very small if you don't have a lot of email on the flash drive. So when you kill it off, the stub is still left in memory and hence you don't see a big delta change in memory free. And you have to remember, something like 80% or more of the apps for iOS are under 30 megs. Only large apps with very large data sets (IE games and certain productivity apps) will show the FULL behavior of iOS.

 

[eddieg]

So there you go the switcher DOES clear RAM!

 

[Skull One]

So there you go the switcher DOES clear RAM!

It has the "potential", yes. That is why I have never stated this as a "clear cut" fact. Because there are variables involved in this discussion that have to be taken into consideration.

But my posts do show the "extreme" condition that iOS can and does handle.

 

[zig9449]

Here is a quick and easy test.

See post #42 for the explination

That's what I been sayin'. That's all. Exactly what those screen shots say.

 

[zig9449]

Grrr.... iPF on the iPhone post pictures differently than on the iPad. Lesson learned today.

But beyond that lets look at this factually.

The middle picture show 55.6% RAM FREE. That is the starting point right after a reboot and giving the phone 1 minute to settle into its normal mode of operation.

The first picture is after I launched Tiger Woods 2012 and got the point that I was at the tee ready to swing. Notice the RAM FREE is now down to 28.7%

The third picture is after I task killed it. Notice the FREE RAM is now 49.8%.

Now for the WHYS...

Every program has two major sections to it.

1) Code
2) Data

The code can be in one of three states:
1) Wired
2) Active
3) Inactive

The data can be in one of four states:

1) Wired
2) Active
3) Inactive
4) Other

And this is all dependent on how the developer codes his app.

Now you will notice there is a difference between the STARTING free and ENDING free. The reason for this is rather simple. The app leaves part of itself in memory for faster launching. This is sometimes called a "Stub". For now we will use that term. The Stub when the program becomes paused moves from wired/active to inactive. Once the Stub is no longer needed it moves from inactive to other or is released.

So the next question is what happened to the DATA.

Easy, iOS didn't see a need to keep it in memory at all because the USER killed the app. The data is dynamic. Which means even if the Stub is still in memory the data isn't worth anything since the next launch will change it. Simply put that is just good Operating System design.

Hope that clears up part of what everyone is seeing.

That is my exact understanding of it. I just couldn't word it with the intricacy such as yourself. I couldn't remember and so I just gave up on trying to explain it.lol.

 

[pianoman]

Wow. This is where one of my posts can be taken EITHER way for the argument at hand. And that is because of "how things work" in iOS.

Technically what I said was that the Task Bar can be used to kill an app. IE force it to reload from scratch. And that the task bar survives a reboot because it is technically a list of last used apps. BUT...

What it leaves in memory when you do that, that is entirely different subject and one I should have been a lot more evidential about in my last major discussion on the subject. For that I apologize.

If you read my post #42 in reference to the pictures I posted on #39 you will see beyond a shadow of a doubt that the task bar can be used to free memory (IE I ain't jailbroke, so I can only kill an app that way).

BUT the issue is "which apps" when you kill them show this behavior. IE the Mail or Message apps won't show this. Tiger Woods 2012 on the other hand does in fact show what iOS does when the task is removed from the Task Bar. The reason for the difference is the amount of memory Mail takes up is very small if you don't have a lot of email on the flash drive. So when you kill it off, the stub is still left in memory and hence you don't see a big delta change in memory free. And you have to remember, something like 80% or more of the apps for iOS are under 30 megs. Only large apps with very large data sets (IE games and certain productivity apps) will show the FULL behavior of iOS.

Well ok. None of Apple's materials point to this behavior as a role of the switcher. And countless of my own tests show no change in RAM at all after releasing an app in the switcher. Maybe there are a couple of apps that can be released using the tech flow you outlined - but it's not an inherent function (by design) of the switcher - again for reasons you outline. I see countless iPhone users clearing out the switcher to try an regain RAM but it really doesn't achieve their goal. I've had two iOS experts verify this for me and I consider you an expert too, Skull. So I'm not sure why you experts disagree. My main point and you're backing me here is that releasing RAM is not an inherent design quality of the switcher. It's like a 1 in 100 chance that the app you choose to release from the switcher will actually get you RAM back. In short it's a waste of time to release apps from the switcher.

In fairness you say "has the potential to" - but the potential is not a design feature at all. Another proof of this is very testable. Load up 20-30 apps in the switcher. Check RAM left. Then start releasing them one by one. The RAM read out in any one of several jailbroken RAM usage apps may go down a little here and there - but when you've emptied the whole switcher you'll never be anywhere near the RAM readout of after a reboot. It's totally random. And any RAM you regain could just as easily be from a timeout of a paused app in RAM releasing on its own.

I'm now quite tired and bored with this topic. :-/... Release apps from the switcher if you wish. But you're really wasting your time - and any Apple Genius at a store will tell you that too along with several developers I've chatted with. I'm gonna still call it a "myth".

 

[zig9449]

Well ok. None of Apple's materials point to this behavior as a role of the switcher. And countless of my own tests show no change in RAM at all after releasing an app in the switcher. Maybe there are a couple of apps that can be released using the tech flow you outlined - but it's not an inherent function (by design) of the switcher - again for reasons you outline. I see countless iPhone users clearing out the switcher to try an regain RAM but it really doesn't achieve their goal. I've had two iOS experts verify this for me and I consider you an expert too, Skull. So I'm not sure why you experts disagree. My main point and you're backing me here is that releasing RAM is not an inherent design quality of the switcher. It's like a 1 in 100 chance that the app you choose to release from the switcher will actually get you RAM back. In short it's a waste of time to release apps from the switcher.

In fairness you say "has the potential to" - but the potential is not a design feature at all. Another proof of this is very testable. Load up 20-30 apps in the switcher. Check RAM left. Then start releasing them one by one. The RAM read our may go down a little here an there - but when you've emptied the whole switcher you'll never be anywhere near the RAM readout of after a reboot. It's totally random. And any RAM you regain could just as easily be from a timeout of a paused app in RAM releasing.

I'm now quite tired and bored with this topic. :-/... Release apps from the switcher if you wish. But you're really wasting your time - and an Apple Genius at a store will tell you that too not to mention developers, in general.

.....................Wow.

 

[pianoman]

.....................Wow.

............... ;-)

 

[Skull One]

Well ok. None of Apple's materials point to this behavior as a role of the switcher. And countless of my own tests show no change in RAM at all after releasing an app in the switcher. Maybe there are a couple of apps that can be released using the tech flow you outlined - but it's not an inherent function (by design) of the switcher - again for reasons you outline. I see countless iPhone users clearing out the switcher to try an regain RAM but it really doesn't achieve their goal. I've had two iOS experts verify this for me and I consider you an expert too, Skull. So I'm not sure why you experts disagree. My main point and you're backing me here is that releasing RAM is not an inherent design quality of the switcher. It's like a 1 in 100 chance that the app you choose to release from the switcher will actually get you RAM back. In short it's a waste of time to release apps from the switcher.

In fairness you say "has the potential to" - but the potential is not a design feature at all. Another proof of this is very testable. Load up 20-30 apps in the switcher. Check RAM left. Then start releasing them one by one. The RAM read out in any one of several jailbroken RAM usage apps may go down a little here and there - but when you've emptied the whole switcher you'll never be anywhere near the RAM readout of after a reboot. It's totally random. And any RAM you regain could just as easily be from a timeout of a paused app in RAM releasing on its own.

I'm now quite tired and bored with this topic. :-/... Release apps from the switcher if you wish. But you're really wasting your time - and any Apple Genius at a store will tell you that too along with several developers I've chatted with. I'm gonna still call it a "myth".

First off, lets get a few things straight for the record.

1) Task killers are a WASTE OF TIME in a properly written OS. Android didn't even reach that level till 2.1 and still has issues. iOS actually took a step backwards with iOS 5.

2) Your experts (I normally NEVER do this out of professional courtesy) are IDIOTS. The task switcher has 3 functions.
a) Show you the last X (usually 80) apps used.
b) Maintain that list thru reboots.
c) Kill a task if the task is still in memory.

Those are facts that can not be argued with. All three items I just listed are DESIGN FEATURES.

3) I can list TWO different scenarios where I can both PROVE and DISPROVE both sides of this coin. BUT only one of them is FACTUALLY correct. And here is why:

A) Remove ALL items from the task switcher, reboot, load Tiger Woods, kill Tiger Woods. OH, I have pictures to back that up on post #39.

B) Remove ALL items from the task switcher, reboot, load ANY 20 apps, load Tiger Woods, kill every app EXCEPT Tiger Woods.

Guess which one show what really happens and guess which one MASKS the truth?


Personally I don't care what you want to believe. It is like asking if humans were created by God or Evolution. It doesn't matter how much evidence you provide, someone is going to be called a liar.

But when it comes to computers it is CLEAR CUT and PROVABLE. I proved it, you either can accept that you are wrong or you don't.

I didn't want this conversation to turn into a "he said, she said" because I just listed the EXACT scenario that causes people to see "their side of the truth". When the reality is, IF you kill an app in the task tray AND it is in memory it will release ALL or PART of the memory back to the OS based on several factors.

I may not be able to get my hands on the iOS' source code, but it is a *nix based OS. And I have studied the *nix OS since Xenix was introduced to the unwashed masses. iOS isn't special. It doesn't do anything revolutionary. It simply does what it was coded to do. And one of those features is killing a task and freeing the memory associated with it beyond the base stub of the application until such time iOS needs to free it to load a new app or expand the data footprint of an existing app.

 

[zig9449]

First off, lets get a few things straight for the record.

1) Task killers are a WASTE OF TIME in a properly written OS. Android didn't even reach that level till 2.1 and still has issues. iOS actually took a step backwards with iOS 5.

2) Your experts (I normally NEVER do this out of professional courtesy) are IDIOTS. The task switcher has 3 functions.
a) Show you the last X (usually 80) apps used.
b) Maintain that list thru reboots.
c) Kill a task if the task is still in memory.

Those are facts that can not be argued with. All three items I just listed are DESIGN FEATURES.

3) I can list TWO different scenarios where I can both PROVE and DISPROVE both sides of this coin. BUT only one of them is FACTUALLY correct. And here is why:

A) Remove ALL items from the task switcher, reboot, load Tiger Woods, kill Tiger Woods. OH, I have pictures to back that up on post #39.

B) Remove ALL items from the task switcher, reboot, load ANY 20 apps, load Tiger Woods, kill every app EXCEPT Tiger Woods.

Guess which one show what really happens and guess which one MASKS the truth?

Personally I don't care what you want to believe. It is like asking if humans were created by God or Evolution. It doesn't matter how much evidence you provide, someone is going to be called a liar.

But when it comes to computers it is CLEAR CUT and PROVABLE. I proved it, you either can accept that you are wrong or you don't.

I didn't want this conversation to turn into a "he said, she said" because I just listed the EXACT scenario that causes people to see "their side of the truth". When the reality is, IF you kill an app in the task tray AND it is in memory it will release ALL or PART of the memory back to the OS based on several factors.

I may not be able to get my hands on the iOS' source code, but it is a *nix based OS. And I have studied the *nix OS since Xenix was introduced to the unwashed masses. iOS isn't special. It doesn't do anything revolutionary. It simply does what it was coded to do. And one of those features is killing a task and freeing the memory associated with it beyond the base stub of the application until such time iOS needs to free it to load a new app or expand the data footprint of an existing app.

I bet he still finds a way to disagree.lol.

 

[eddieg]

He will its been proven that he is wrong so he can say whatever he wants!

 

[pianoman]

He will its been proven that he is wrong so he can say whatever he wants!

Remember he said it only works for SOME apps in the switcher and that it's a waste of time to kill apps from the switcher. You two can accept your expert with his proof. I will accept my experts and my own scores of tests I've done with the switcher that prove easily that for 99.9 % of apps "I use" you get back no RAM. I dont know why Skull and Eddie are getting different results than I am. I do not own Tiger Woods.

I will accept Skull's input that for that one app the app switcher releases RAM. His pictures are proof of that. I'm not an idiot. I've been working with iOS as long or longer than anyone here on the forum. I will not be made fun of or called an idiot. And it's against forum rules anyway. This is a complicated topic surrounded by much myth. The app switcher is a glorified set of shortcuts BY DESIGN. Apple says so in their literature. End of story "for me". Kill all the apps you want believing you'll get all your RAM back. No problem here.

But Eddie I accept you see what you're seeing. And Zig I accept your general explanation. And thanks, Skull, for your clear explanations and testing for us! On the important points (waste of time to clear apps from the switcher, not all apps will free RAM upon releasing them from the switcher) we are in agreement. On Apple's design for the switcher we'll have to agree to disagree.

Good discussion.

 

[LaurenRae]

A related question: does the cache of an iPhone even need to be cleared?

I have heard (rightly or wrongly) that with Windows products, when you close programs, they leave pieces of themselves still open, so eventually, all those open program pieces will clog the memory (or something) and make the computer run more slowly. This appears to be true, since with all the Windows-operated computers I have owned, I have had to restart them periodically, believing I was closing down all programs in the process and starting over. I know they computers HAVE operated better after having restarted them.

I have heard (again rightly or wrongly) that Apple products don't do that, that you can leave programs open or close them and they don't cause problems. Thus, using a whole bunch of programs - or apps - shouldn't mess up an Apple computer or iPhone.

Is there any truth to this?

It had appeared through much of the discussion that occurred after I posted this question that most, if not all, apps do close when they are not being used. I have noted that some apps - such as when I stream from a radio - will continue to play music when I am doing something else on the phone - playing a game, for instance.

Then, there was a post that said that parts of an app do stay open in order to make it easier to open them again later. Apparently, Windows does this to an obnoxious extent, where I have to restart to cancel out any problems (and it didn't work this morning, by the way - I am having difficulty posting this from my laptop running Windows 7 - it keeps not responding). From what I've heard, Apple products don't do this, at least not to the extent Windows does.

Any thoughts?