1. Are you ready for the Galaxy S20? Here is everything we know so far!
Thread Status:
Not open for further replies.

Why you don't need a task killer

Discussion in 'Android Apps & Games' started by ardchoille, May 13, 2011.

  1. HankAtrix

    HankAtrix Well-Known Member

    It looks like I need to repeat a couple of themes from this post (from me). I have never said Android needs a task killer to run or that it is a cure-all (panacea) for Android. But it does do a better job of killing apps than doing it yourself or simply letting Android do it alone.

    As most people know, third party apps are a way of life for Android. You can improve almost every capability with these applications. If you have a lot of applications, nn this case ATK does a better job of killing tasks than doing it alone or letting Android do it. For now, it is an effective option. It is much more efficient.

    I am arguing that Android will run faster when you have most of the apps closed. Many people seek that ability to make their phones faster and improve battery life. ATK does both of these extremely well. The fast that it continues to be a top download on the Market would suggest that many/most phones users are not satisfied with Android alone.

    There may be other task killers that cause more problems (and have had user have bad experiences). The only one I’ve used is ATK.

    Like any 3rd party apps, ATK isn’t for everyone.

    I’ve used benchmarks to demonstrate to the increases (2 different benchmarks).

    On the Atrix, the Task Manager doesn’t show Cached apps (but it does show running apps and amount of CPU used by each). But this is where you can set up the phone’s own task killer (the phone calls them the “Auto-End” list).

    With regards to the Anti Task Killer contingency, there is a notion that the use of task killers is not beneficial. Like any debate, there are many sides of the argument and strong opinions.

    I hope this helps!
     



    1. Download the Forums for Android™ app!


      Download

       
  2. elemteacher2be

    elemteacher2be Well-Known Member

    Okay - I'm good on space.

    Here's another (probably dumb) question - what is the difference between USB storage and internal storage?
     
  3. sharkerbaby

    sharkerbaby Member

    No it doesn't mean that at all. No one is able to ascertain the reason for the download, nor are you able to determine how many of those people no longer use the app. You can count me among them. Yes as a new android user, I believed as you do and so yes one of the first apps i downloaded was atk and went along believing it was helping my device and allowing me to use it's resources with a high level of efficiency. Then i began educating myself and reading and lo and behold discovered i was actually quite ignorant. The most convincing testament was when i read an article/posting written by the android os developers themselves arguing against task killers - i ask you... who would know better the code writers or an end user?

    I have had my device for almost 2 yrs. It's an original Droid and, as we all know, compared w/ newer devices, it's specs are not even close. I have over 120 apps installed and I use it constantly for just about everything imaginable. I uninstalled atk well over 1 1/2 yrs ago, my device is as fast and responsive as i could ever ask for. Once it is booted and the SD card mounted, there is virtually no delay in any task i ask it to do. On the few occasions when i have experienced lag it has always been either 1) a configuration change that i made - user error or 2) a rogue app that is not written properly and refuses to either go idle or be cached and invariably finds itself uninstalled
     
    marinegundr and ardchoille like this.
  4. ardchoille

    ardchoille Android Expert
    Thread Starter

    Usta is correct:

    • Android is hard coded to automatically kill a task when more memory is needed.
    • Android is hard coded to automatically kill a task when it
     
  5. elemteacher2be

    elemteacher2be Well-Known Member

     
  6. HankAtrix

    HankAtrix Well-Known Member

    Dave - that is the point exactly. phones are not parallel (many processors) it it just multi-tasking. I am saying that if it was parallel hardware, it would continue to perform with little degradation when you add apps... but since it isn't, it has degredation of performance with many apps.
     
  7. HankAtrix

    HankAtrix Well-Known Member

     
  8. dvhttn

    dvhttn Android Expert

    The phones I mentioned are dual-core, therefore they multi-process, as well as multi-task obviously. As for degradation of performance I think this has been covered. People don't see performance degrading. You're flogging a dead horse here I'm afraid.

    dave
     
  9. HankAtrix

    HankAtrix Well-Known Member

    How can you say that it isn't in demand by many, many people? ATK alone has been downloaded over 280,000 times. Like any apps, many are not using it. That doesn't take away from the need for performance and battery-life improvement.

    The fact that you don't need it or choose to use it doesn't change the impact that ATK has had on my phone and hudreds of thousands others. In your case a configuration change helped. In my case ATK has made the phone notably faster (and I use it to efficently close apps).

    Have a great day!
     
  10. HankAtrix

    HankAtrix Well-Known Member

    That is 1 persons opinion
     
  11. elemteacher2be

    elemteacher2be Well-Known Member

    Question about the WatchDog App. I installed it while reading this thread and was wondering if the "Memory (free/total)" in the stats part is the internal memory that I should be looking at for that 30% mentioned earlier?
     
    blahsaysmeu2u likes this.
  12. Usta

    Usta Android Expert

    Ardchoille elaborated further based on my post. What he said above confirms also my beliefs. I've stopped using task killers very long time ago, and since then never needed one.

    This is year 2011, and I expect a smartphone OS to manage its own resources without my intervention. In my opinion, Android OS (especially with Gingerbread) does that job pretty well.
     
    ardchoille likes this.
  13. HankAtrix

    HankAtrix Well-Known Member

    We expect a lot out of our phones. The fact that ATK does a more efficient job of closing apps (I can't speak to other task killers). The fact that Android can handle it doesn't mean that there aren't better ways to do app closing. Have you downloaded any apps that improve your phones abilities? I have downloaded dozens for things Android can do alone (calendar, keyboard, etc). In my case ATK is a dramatic improvement.
     
  14. Usta

    Usta Android Expert

    Well, if you like so much ATK, then it's surely your choice.

    I like to try new apps and my phone has always 100+ downloaded apps installed on it. At one time I came even close to having about 200 user apps on my phone, when I was testing what the best apps are for me in different categories. By the way, these are extra to whatever pre-installed apps come on system partition.

    Bottom line: yes I try many apps on my phone, but I still don't see any need to use a task killer to close them.
    Now, if that was on any other OS, like Symbian, iOS or Windows Mobile, a task killer/manager would probably be standard in my apps drawer.
     
  15. HankAtrix

    HankAtrix Well-Known Member

    Their premise is that they improve performance and battery life. Their premise is accurate. If you don't need these, then using a task killer won't fit for you. Keep in mind that all you have to do is use the ignore list for apps that you don't want to close.
     
  16. thewhopper

    thewhopper Newbie

    I'll say this...

    Although I'm learning Java and Android programming right now, I don't fully understand the internal workings and concept as to what Task Killer does or the pros and cons as to what it does. Nor do I have the time to delve deep into further.

    However, I will tell you this.....

    When my HTC Eris starts lagging, I run Task Killer.....and it works.

    If you're rooting your phone and spending hours customizing your phone, customizing your apps, removing bloatware, etc...then you 'might' not need task killer.

    ...but for most people (I fall in this category even though my ERIS is rooted), the Task Killer works, it's quick and easy, plus it saves me time. :cool:
     
    HankAtrix likes this.
  17. lunatic59

    lunatic59 Moderati ergo sum
    Moderator

    Again we see the logical fallacy of "begging the question" rear it's ugly head. For this statement to be valid, we must assume that closing apps efficiently using ATK is a fact. Yet there is no factual evidence to back it up. If we assume speed is equal to efficiency, then we are furthering the assumption by committing a second logical fallacy.

    Here we have a false conclusion. It is stated as:
    1. Android has built in functions.
    2. Functions can be improved with third party applications.

    Therefore, because Android can kill processes and ATK is a third party application, ATK improves the functionality of Android.

    This false conclusion is followed by a compositional fallacy (also seen as improper set theory) where a characteristic of a whole class is drawn from a subclass. Because we have a sample set of apps that improve functionality over the stock apps, we are to assume that apps in general improve functionality.

    And we continue with the fallacy of observational selection. As explained earlier, it is only taking any benefit into consideration without regard to and negative impact or consequences.

    Let me try and theorize why we may be seeing a performance increase with the use of a task killer, and how it is pertinent to observational selection.

    As we all should know by now, Android utilizes free memory -- memory not used by actual running process -- to cache recently used apps in state and cache frequently used apps in anticipation of their use. If a process ends and is removed from memory, Android will attempt to load a process or app based upon a prioritization algorithm. If more room is needed for a higher priority app or service, processes will be safely ended in order of ascending priority.

    So now we have our current dilemma -- we have 30 apps and services listed as running and wish to improve performance of Android. Does Killing 20 of those listed apps and services actually improve the performance?

    The answer may be yes, for the short term. Android is designed, like any modern OS, to prioritize tasks. User spawned tasks always get the highest priority, so that when you tap a button or update a widget, the user is not asked to wait for a system process and giving the appearance of a lagging device. Since the launching of a task killer is essentially a user spawned process, it will take priority from Android's task manager. And, while you continue to interact with the screen, those actions will be given priority.

    Now, during the period immediately following the use of a task killer, if you were to launch another app or run a benchmark (assuming that those apps or benchmarks don't require any of the killed processes which would need to be respawned immediately) because of the temporary lack of cached apps you may get a false sense of efficiency and inaccurate benchmarks. However, as soon as there is even the briefest period of inactivity, Android's task manger will begin to reclaim the unused memory, consuming resources and power that would not have been necessary otherwise.

    It explains why after killing 20-30 applications or processes and using 5 to 10 applications it is necessary after a certain period of time to kill 20-30 again.
    Not taking into account the time it takes to setup a task killer or the resources itself consumes just to exist as an app or widget, if the immediate temporary performance boost is not enough to offset the perceived lag while Android would have normally managed its own processes then the efficiency is zero.

    What I suggest is that a short term temporary performance boost is indeed possible; that doing this does not increase the phone's efficiency; that there is an ultimate power and resource cost to forcing the OS to perform tasks that would have otherwise been unnecessary and that prematurely and abruptly stopping groups of apps or processes can be detrimental to the overall operation of the device.

    Weighing the minimal temporary benefit to the lowered efficiency and potential negative consequences, the use of an automated task killer is not recommended, and unnecessary.
     
    Buddha64 likes this.
  18. HankAtrix

    HankAtrix Well-Known Member

    It is a fact that Android runs faster with fewer apps running. It will take running a few apps to see a discernible difference in the running apps by closing apps. Android is multitasking but it is not parallel (as you run apps – whether 1 or 20) it has to share the single CPU among the running applications and processes. We only have 1 or 2 CPUs in each phone (it is the CPU and memory that is the constraint not Android itself).


    This is not just a short term benefit. Over any period (short, medium, long) your phone will run faster with fewer apps running. This is not a short term fix, it can (and is in my case and most users case) a part of my daily routine. I like ATK because it does app closing with many options (no kill list, running a standard app close every hour and as a widget on my home-screen).


    This process takes less time than other ways I’ve tried (ie. Closing apps one at a time or using TaskManager) which leads to better battery life (less time on the phone and closing battery intensive apps). ATK is simple and effective.

    What else do we need to do to prove it? Personal experience, benchmarks, hundreds of thousands of users, 3500 excellent ratings on Android Market. And task killers are being added to Android itself (and are already available as standard on the Motorola Atix).

    There are no negative consequences if you use an ignore list. Who cares if a few apps need to be re-opened? A great majority (in my case 30 to 1) can be closed. I don’t mind if a few need to be re-opened by me or Android, the ratio works. If you want to not interfere with other apps, add them to the ignore list). As I’ve said, ATK gives you a recommended list of processers and apps and I’ve added Ultimate keyboard, calendar, Lookout, and JuiceDefender.

    I hope this helps
     
  19. lunatic59

    lunatic59 Moderati ergo sum
    Moderator

    It is a fact that the more processes running concurrently, the more of a load it will place on any system. However what is NOT a fact, and one that is continually dismissed is that most apps listed as running are not actually consuming any cpu cycles and are only running in the sense that they are cached in memory. If you kill any app that is actually running you run the risk of data corruption. If you doubt this, why not do a test on one of your database systems ... while it's doing a disk write, yank the plug.

    This is ONLY true if the apps are actually running, as in consuming resources and CPU cycles. They are not. Closing apps on a regularly scheduled basis, especially when there is little or no phone use is counterproductive and wasteful.

    This is a non-issue. Again, begging the question fallacy. These apps are not consuming CPU cycles so if there are 20 or 50 of them cached in state there is ZERO impact on the processor.

    Not using a task killer and letting Android do what it was designed to do take NO time and is more efficient. Battery intensive apps only put a load on the battery when consuming CPU cycles. If you kill an app while it's actually using the CPU, you run the risk of file corruption.

    How about actual proof bases on technical facts from qualified sources instead of repeating the same fallacious arguments over and over?

    There are no negative consequences if you avoid task killers altogether. Why risk problems for questionable benefits?
     
    ardchoille likes this.
  20. Slug

    Slug Check six!
    VIP Member

    Well, let's see what one of the Honeycomb devs thinks about task killers here. Or the viewpoint of another Android software engineer in this article? When the folks who write the actual code assert something, it tends to based on solid fact.

    There are also many detailed articles from the likes of Lifehacker and InfoWorld, to name two that I have bookmarked, which back this up.

    Nope, it's a Task Manager. The difference may be subtle but it's important.
     
  21. hansschmucker

    hansschmucker Android Enthusiast

    I love this quote:

    “this is not a task we should ask humans to do, because, first of all, humans don’t do a very good job of it. They don’t have enough information and enough context.”

    Which pretty much sums up my feelings towards task killers on Android: To make them useful you need to know a lot more than most people do. If you are an Android Ninja with mystical Android skills they are definitely able to help: But for most people there's no point since they are unable to get this kind of information. Plus, it means a whole lot of research and at least I don't have the time for that.

    Using a task killer is not a sin... but using one with the default settings is.
     
  22. KingOfGreen

    KingOfGreen Android Expert

    oh HankAtrtix, whats wrong with you! you just keep talkin the same BS. Why dont you show us some articles from devs explaining why you should be using a task killer???? cuz there are plenty for the contrary! you said
    now just because your company build unix systems doesnt mean you know a thing about how unix/linux/android works, ESPECIALLY more so than the people who WROTE IT THEMSELVES!!! like i said, provide some sources of people who have a good rep. around the android parts saying you should be using a task killer.
     
  23. hansschmucker

    hansschmucker Android Enthusiast

    Don't you think you're all making this a bit too personal... As much as I have an opinion about this, I still think it's nothing worth fighting about.
     
  24. lunatic59

    lunatic59 Moderati ergo sum
    Moderator

    It has been a lively discussion, but it does seem that a great deal of effort is being made to keep it on topic.

    But let's just say a woman walks into a childbirth support group and begins advocating the use of alcohol as an effective pain killer for the discomfort of pregnancy. (Yes, I know it's an extreme analogy) On the other side you have a midwife who, although is not a doctor, is backed by medical evidence and scientific data to the contrary. In the room are a dozen or so pregnant women all in varying degrees of discomfort looking for answers.

    The advocate of the questionable practice cites the fact that she drank while pregnant and felt great and all her children turned out okay. The midwife on the other hand cites science and shows the dangers but can only offer the women modest relief through proven practices.

    Even though it is possible that all 12 women could drink through their pregnancies and deliver healthy babies while feeling pretty good, would it not be in the best interest of all parties to argue against the possible dangers?

    When it comes to task killers there is little benefit and potential for harm from dwindling battery life to corrupt data to the possibility of a semi-bricked phone in need of a factory reset. I believe it is irresponsible to argue in favor of such practices and equally irresponsible to not dissuade users from the same.
     
    Buddha64 and ardchoille like this.
Loading...
Similar Threads - Why don't need
  1. Drchiefd113
    Replies:
    5
    Views:
    482
  2. kumaranil13k
    Replies:
    0
    Views:
    479
  3. kumaranil13k
    Replies:
    0
    Views:
    285
  4. dorlow
    Replies:
    3
    Views:
    779
  5. JimmyPixel3a
    Replies:
    21
    Views:
    1,377
  6. iliji
    Replies:
    0
    Views:
    907
  7. dontpanicbobby
    Replies:
    46
    Views:
    1,426
  8. FensterBaron29
    Replies:
    5
    Views:
    7,534
  9. User123app
    Replies:
    0
    Views:
    709
  10. HeartOnFire69
    Replies:
    2
    Views:
    872

Share This Page

Loading...