Common Android Myths/Explanations

[YankInDaSouth]

OK - thanks. I must have missed it in there. Will Check them out ...

 

[cvpcs]

Don't know if this is the right thread for this ... I know you looked at FRG22 but will this latest news have you reexamine it and potentially add an update to 1.0? Not sure it's needed cuz this things a rockin' for me :tongue:

The information in the myth page is absolutely true, and I have confirmed it as best I can. That being said for now I won't be including FRG22 release in a rapid update to Sapphire as of now, as it isn't important.

In the next release I will incorporate it to stay current, but unless I need to change the build props to reference FRG22 now, there won't be an update to modify those props.

 

[YankInDaSouth]

Don't know if this is the right thread for this ... I know you looked at FRG22 but will this latest news have you reexamine it and potentially add an update to 1.0? Not sure it's needed cuz this things a rockin' for me :tongue:

The information in the myth page is absolutely true, and I have confirmed it as best I can. That being said for now I won't be including FRG22 release in a rapid update to Sapphire as of now, as it isn't important.

In the next release I will incorporate it to stay current, but unless I need to change the build props to reference FRG22 now, there won't be an update to modify those props.

Got it! Thanks so much for taking the time to reply and for such excellent work!! I now have found a ROM to call home!

 

[silk]

First off, that was a mistake made based on some misinformation that I had received that I believed to be reliable. I have seen your post and have updated my other thread. Also, you will notice that this page does not mention it either.

I understand your concern with my other thread sparking rumors and I apologize for the misconception but I don't really appreciate your tone as it is essentially implying that this post has no value.

While I admit that my post does not list sources such as wikipedia it is because a lot of the sources are myself. That being said, I have always been open to people proving me wrong and will gladly update anything that is proven to be incorrect, but please don't misunderstand my intentions.

Hey man, we all make mistakes. Don't sweat it. I saw the same thing back at the end of June and then saw something at the beginning of July saying otherwise. That was all over a month ago and something new may have come out recently. That's why I wanted a source to go with what you originally posted. And bringing it up here, seemed appropriate because unsubstantiated rumors are made of the same thing myths are, and that's what this thread is about, no?

 

[Martin030908]

cvpcs... great info.

 

[cvpcs]

First off, that was a mistake made based on some misinformation that I had received that I believed to be reliable. I have seen your post and have updated my other thread. Also, you will notice that this page does not mention it either.

I understand your concern with my other thread sparking rumors and I apologize for the misconception but I don't really appreciate your tone as it is essentially implying that this post has no value.

While I admit that my post does not list sources such as wikipedia it is because a lot of the sources are myself. That being said, I have always been open to people proving me wrong and will gladly update anything that is proven to be incorrect, but please don't misunderstand my intentions.

Hey man, we all make mistakes. Don't sweat it. I saw the same thing back at the end of June and then saw something at the beginning of July saying otherwise. That was all over a month ago and something new may have come out recently. That's why I wanted a source to go with what you originally posted. And bringing it up here, seemed appropriate because unsubstantiated rumors are made of the same thing myths are, and that's what this thread is about, no?

Agreed. Hopefully that wiki page will remain updated with only truthful information (or in the absence of any proof logical conjecture). It (and this thread) are intended to be a place where people can find some meaning to the madness

 

[captdroid]

Thanks for sharing. For the most part, I've seen these addressed before, but somehow these myths continue to persist. Nice to see them all in one place. The worst myth, IMO, is that of task killers and keeping MB free. The problem is that most people are accustom to Windows desktops, where free RAM directly correlates to the performance of whatever program you're looking to open next. Unfortunately, as CVPCS states, keeping memory free is "...a horrible idea and just leads to problems with redraws and instability." This is exactly why I abandoned BB long ago. It had that damn 50MB free at all times script, which meant that if I was using navigation and Pandora at the same time, Pandora would periodically get killed because Navigation needed a bit more memory. The g/f tried to use the browser (nav and Pandora were already running) one time on a trip to look for restaurants along our path...yeah, that didn't go well because the script caused programs to keep getting killed.

Hopefully this information continues to spread around, but with 200,000 new Android activations a day, most of whom I'm assuming are not looking toward the dev community, it'll be a long battle.

You know when I got my DROID one of the first things the Verizon rep said in a phone conversation was I should get Advanced Task Killer and that it really works well. Knowing that Verizon reps generally work from some sort of scripts it's no surprise people are getting bad advise. :icon_eek:

 

[aminaked]

Another big thing is that the default email client (at least on original droid) is disastrously bad. One bug has it attaching a message at random when replying to an email.

I spend about an hour/day on droidforums telling people not to use that crap.

 

[PJnc284]

You know when I got my DROID one of the first things the Verizon rep said in a phone conversation was I should get Advanced Task Killer and that it really works well. Knowing that Verizon reps generally work from some sort of scripts it's no surprise people are getting bad advise. :icon_eek:

I've tried my damndest over on the Verizon Community forums to convince people that they don't "need" task killers but even some of the mods over there are pushing task killers and the users can't figure out why they end up having so many problems.

 

[banderbe]

I did a quick write up to demystify kernel "development".

The link is in my sig.

 

[banderbe]

You know when I got my DROID one of the first things the Verizon rep said in a phone conversation was I should get Advanced Task Killer and that it really works well. Knowing that Verizon reps generally work from some sort of scripts it's no surprise people are getting bad advise. :icon_eek:

Once in a while if my Droid gets sluggish I'll use ATK and it perks right up. I wouldn't use it regularly for no reason though. The idea that Android memory management is flawless and never needs intervention is silly.

 

[banderbe]

From the wiki:

" Android is built as an OS to run with its memory full at all times, because that allows it to run faster."

I'll call bulls**t. I agree it's built to run with its memory full. I'm quite familiar with how it manages activities on its stack.

I disagree that it runs faster that way. Android runs SLOWER once it has to start saving application states, killing the activities, etc. Not much slower, and certainly not perceptible to the user but when only a few activities are running Android doesn't need to start doing what is basically a kind of paging.

 

[captdroid]

You know when I got my DROID one of the first things the Verizon rep said in a phone conversation was I should get Advanced Task Killer and that it really works well. Knowing that Verizon reps generally work from some sort of scripts it's no surprise people are getting bad advise. :icon_eek:

Once in a while if my Droid gets sluggish I'll use ATK and it perks right up. I wouldn't use it regularly for no reason though. The idea that Android memory management is flawless and never needs intervention is silly.

I think the point is that using Task Killer programs generally are a bad idea for most people.

CVPCS IMO says it best here

 

[TuBitMittens]

Nice info. Personally, I couldn't give a flying fish about

​

benchmark scores, let alone I still don't truly know what they really are. As long as my phone is performing good at a decent speed and I get decent battery(battery has been hard on me lately) then I'm fine.
Also another thing I don't quite understand; I do happen to have system task panel app, which monitors my phone's performance and can also act as a task killer(although I'm trying to figure out if not monitoring my phone with it makes battery any better)ANYWAY, what I don't get is, aside from those internal apps that are suppose to be running at all times, why does it sometimes show my apps that are running that the native task killer doesn't?? Is that just cached memory info of apps I was previously running or something?? I don't really know.
I always make sure to completely close apps by using the back button.

 

[cvpcs]

From the wiki:

" Android is built as an OS to run with its memory full at all times, because that allows it to run faster."

I'll call bulls**t. I agree it's built to run with its memory full. I'm quite familiar with how it manages activities on its stack.

I disagree that it runs faster that way. Android runs SLOWER once it has to start saving application states, killing the activities, etc. Not much slower, and certainly not perceptible to the user but when only a few activities are running Android doesn't need to start doing what is basically a kind of paging.

First off let me say there's no need to call bull**** as I'm not attempting to fool anybody with that wiki. Its enough to simply state that you disagree.

As for your comment, you bring up an interesting point. In the event that the apps running in the background are actually using cpu cycles by performing tasks, you would be correct that this multiprogram approash would cause changing of states that have to be loaded through the cpu to run those processes. However, this is unlikely to cause a significant performance hit as this is done all the time anyway between all of the system processes, plus RAM is extremely fast.

My statement that using all of the ram is faster comes from the simple fact that even though nand is fast, ram is faster, especially on some devices that use the sd card to store apps, loading an app back into memory will take proportionally longer than switching a state back and running it in memory. if the app is already in memory you remove the need for the whole loading step.

This is my reasoning for declaring it faster, however I do not have any benchmarks to back it up so perhaps I will reword my assumption and post the information for others to evaluate.

Also, thanks for the post on kernels, I completely forgot about that one!