The Windows 10 April 2018 update has arrived! Get the new Dell XPS 15, starting at $999.99
04-28-2015 10:12 AM
43 12
tools
  1. D M C's Avatar
    What does "apps launching later" on WP have to do with "load performance"? Even if we had all the necessary data, I doubt we'd ever find a statistical correlation between those two things.

    What we would find is that WP apps in general all tend to load slower than their Android or iOS counterparts. If we found an app that is completely borked on iOS and entirely tweaked out on WP, we'd maybe be able to demonstrate the opposite situation, but assuming an app with similar features and a similar amount of effort being invested, the WP app will launch slower. Pretty much always.

    I already mentioned why. That's just a result of how WP is designed. Developers certainly do invest less effort for their WP apps. The typical result is complaints about WP apps missing features, but that lack of effort is not the main culprit here.
    It does. Most of the apps launched at beta and later it optimized. Take an example of VLC. Can you tell me how long it will take VLC to went from Beta to Stable version. (it will takes months or may be year) So, till then it will be slow and buggy and VLC for iOS and Android it tends to be perform better than WP. WHY?
    because of the time it was released.
    If VLC launched same time or almost same time on all OS then situation might be different. It will perform same or almost same on every OS.

    You also said "Developers certainly do invest less effort for their WP apps".
    I wonder why? What is demotivating them? is it the market share or what?
    04-21-2015 06:09 PM
  2. a5cent's Avatar
    It does. Most of the apps launched at beta and later it optimized. Take an example of VLC.
    Why many devs invest less effort into their WP apps is kind of off topic and not something I want to get into here.

    You are presenting VLC as the main evidence for your case, by claiming that it loads slower on WP than on iOS, as a result of its beta status. The thing is, VLC will in fact always load slower, even a year from now and after shedding the beta moniker. It may be faster than it is now, but it won't ever load as fast as on iOS. You are free to continue claiming that the VLC team is just failing to do the optimization work they should be, but after testing a larger sample of 3rd party apps (not games), you'd realize that not a single dev is doing that "optimization" work. At what point would you change your mind and say its a systemic problem that isn't just a result of devs not doing their jobs?

    In truth, many WP devs don't put in the effort to optimize load performance, which is when things get really bad, often also coupled with a lot of lengthy resuming. However, there's also a substantial group of devs that does care about load performance. The problem is that the later group must contend with computational processes that just don't exist on iOS. No matter how much those processes are optimized (by MS or third party devs), they will never be as fast as iOS where they don't exist at all. That's the root issue...
    Last edited by a5cent; 04-22-2015 at 05:07 AM.
    04-22-2015 03:49 AM
  3. jack dempsey's Avatar
    I think It's a defect of Windows Phone because apps are at least as capable or more on Android, and they start faster than in Windows Phone. If Windows Phone is doing extra work, that work is not needed, they should remove it, and that's a defect.
    Makm likes this.
    04-24-2015 12:16 AM
  4. Makm's Avatar
    I think that is how android operates, I see often even with 2gb of ram, in idle case, android is consuming 1.6gb, whilst If I compare my phone using the Microsoft tools it shows around 60% amount of ram available.Yes it depend on the various backhand processes but visually that is what I can observe.

    Windows 10 for phones is getting there I guess, It already opens some apps faster compared to win8.1, resuming/loading is gone.
    Had compared whatsapp on a similarly spec android device with my 1020(win8.1), the android was significantly faster, now I did the same test with windows 10 for phones, it is now matching it, ignoring the animation and al
    Conclusion anyone?
    I noticed that wp10 keeps EVERY single apps open at the background. Even all the settings pages, messaging app, phone app etc. Just like facebook used to be open in 8.1. You have to force close everything by long pressing the back button. This is kind of annoying.
    04-24-2015 12:34 AM
  5. raycpl's Avatar
    I noticed that wp10 keeps EVERY single apps open at the background. Even all the settings pages, messaging app, phone app etc. Just like facebook used to be open in 8.1. You have to force close everything by long pressing the back button. This is kind of annoying.
    Agree, but thats W10TP.... this should be history when W10 rolls out...
    04-24-2015 12:38 AM
  6. a5cent's Avatar
    If Windows Phone is doing extra work, that work is not needed, they should remove it, and that's a defect.
    Yeah, right... 😩
    Consider for example the tombstoning process, which can occur when an app is launched.
    Unloading a rarely used app from memory while another is loaded, and doing so before the point where the OS has no choice but to immediately force-close an app due to memory pressure, represents a tradeoff. This sacrifices some load performance, in exchange for NEVER having to expend CPU cycles on force-closing an app while the user is interacting with the device.
    Expending CPU cycles on force-closing an app while a user interacts with the device has the potential to cause stutter/lag on low-end Android hardware. Doing this differently, is how WP achieved its reputation for being smooth and lag free. This mechanism is one reason why WP is competitive in those markets where unsubsidized phones can't cost more than $160 to be successful. It's partly why WP has a shot at success at all.
    How can you or Makm decide if those processes are needed (or are defects) when you have absolutely no understanding of why they exist in the first place?
    If you understood the implications, there'd be a valid debate to be had over whether each tradeoff is worthwhile, but judging them all to be defects based on nothing but ignorance...
    HeyCori likes this.
    04-24-2015 03:12 AM
  7. jack dempsey's Avatar
    Yeah, right...
    Consider for example the tombstoning process, which can occur when an app is launched.
    Unloading a rarely used app from memory while another is loaded, and doing so before the point where the OS has no choice but to immediately force-close an app due to memory pressure, represents a tradeoff. This sacrifices some load performance, in exchange for NEVER having to expend CPU cycles on force-closing an app while the user is interacting with the device.
    Expending CPU cycles on force-closing an app while a user interacts with the device has the potential to cause stutter/lag on low-end Android hardware. Doing this differently, is how WP achieved its reputation for being smooth and lag free. This mechanism is one reason why WP is competitive in those markets where unsubsidized phones can't cost more than $160 to be successful. It's partly why WP has a shot at success at all.
    How can you or Makm decide if those processes are needed (or are defects) when you have absolutely no understanding of why they exist in the first place?
    If you understood the implications, there'd be a valid debate to be had over whether each tradeoff is worthwhile, but judging them all to be defects based on nothing but ignorance...
    Do you have a source to back that up?

    In my experience, stutter in Android is more related to the bad performance of lists. If "Tombstoning" has an impact is minimal. If that's the case then It's a defect of operating system, a bad design, Microsoft should correct this problem in Windows 10.
    04-24-2015 08:49 AM
  8. a5cent's Avatar
    Neither Android nor WP has full control over how long an app takes to shut down, as that's largely defined by the app (cleanup, saving state, etc). Both OSes will force close an app that takes longer than a specified amount of (wall-clock) time, but until that happens, a terminating app is free to use CPU cycles for multiple seconds. As such, your opinion that the impact of tombstoning an app is minimal makes no sense. Multiple seconds is anything but "minimal", and the exact duration differs for every app. Often times it won't matter, but sometimes it does.
    All this is app lifecycle 101 and documented in the developer docs of both platforms.
    You might want to reconsider whether you're qualified to determine what is a defect and what isn't. Either way, I'm signing off here.
    Laura Knotek and HeyCori like this.
    04-24-2015 10:55 AM
  9. jack dempsey's Avatar
    Neither Android nor WP has full control over how long an app takes to shut down, as that's largely defined by the app (cleanup, saving state, etc). Both OSes will force close an app that takes longer than a specified amount of (wall-clock) time, but until that happens, a terminating app is free to use CPU cycles for multiple seconds. As such, your opinion that the impact of tombstoning an app is minimal makes no sense. Multiple seconds is anything but "minimal", and the exact duration differs for every app. Often times it won't matter, but sometimes it does.
    All this is app lifecycle 101 and documented in the developer docs of both platforms.
    You might want to reconsider whether you're qualified to determine what is a defect and what isn't. Either way, I'm signing off here.
    I understand the life cycle of apps, but I didn't find anywhere that WP terminate apps only when starting others and Android at any moment. That's why I asked for the source of that statement.

    The main cause of stutter on Android is poor performance of lists.If I re-start an Android device and only run one app, for example, Play Store, I still have stutter. So, that claim that WP perform better because the timing of app terminations is wrong, that's not the main factor.

    If in WP apps take 10-15 seconds to start, that's an unacceptable defect, and Microsoft has to fix it.
    04-24-2015 11:59 AM
  10. lexus232's Avatar
    I am inclined to think slow and unstable apps, like Spotify, for example, are caused by poor optimisation and coding., rather than the way system itself.
    04-24-2015 06:42 PM
  11. Yazen's Avatar
    I understand the life cycle of apps, but I didn't find anywhere that WP terminate apps only when starting others and Android at any moment. That's why I asked for the source of that statement.

    The main cause of stutter on Android is poor performance of lists.If I re-start an Android device and only run one app, for example, Play Store, I still have stutter. So, that claim that WP perform better because the timing of app terminations is wrong, that's not the main factor.

    If in WP apps take 10-15 seconds to start, that's an unacceptable defect, and Microsoft has to fix it.
    Have you read this yet? https://msdn.microsoft.com/en-us/lib...vs.105%29.aspx

    "If new apps are launched after an app has been made dormant, and these applications requires more memory than is available to provide a good user experience, the operating system will begin to tombstone dormant applications to free up memory."

    512MB RAM pretty much guarantees that the background app will get tombstoned XD

    EDIT: Was not too clear, pretty sure they are not terminated each and every time. That would be bad for performance!
    HeyCori and a5cent like this.
    04-24-2015 06:45 PM
  12. a5cent's Avatar
    ^ You're right. Tombstoneing doesn't happen each and every time. It says apps are tombstoned only if more memory is required than is available. A 512MB device can fit multiple apps into memory, about three on average, so tombstoning is far from guaranteed to occur. Obviously, it occurs far less often on a 1GB device.
    And thanks for the link. I'm not into doing other peoples homework, but that's one place in the docs where tombstoning is temporally associated with the launching of an app.
    04-25-2015 05:14 AM
  13. jack dempsey's Avatar
    "If new apps are launched after an app has been made dormant, and these applications requires more memory than is available to provide a good user experience, the operating system will begin to tombstone dormant applications to free up memory."
    That quote describes the behavior of both operating system, Android and WP. If the system doesn't have enough memory to launch an app, It kills suspended apps. Both systems also kill suspended apps when the active app increase its memory usage.
    So, there is no difference.

    And thanks for the link. I'm not into doing other peoples homework, but that's one place in the docs where tombstoning is temporally associated with the launching of an app.
    When you get caught the best way out is an apology.
    04-25-2015 01:56 PM
  14. Yazen's Avatar
    That quote describes the behavior of both operating system, Android and WP. If the system doesn't have enough memory to launch an app, It kills suspended apps. Both systems also kill suspended apps when the active app increase its memory usage.
    So, there is no difference.
    The quote is ambiguous. What constitutes a "good user experience"?

    In WP8, IE11 would almost always be tombstoned, which was annoying because when the app was reanimated, the address bar would not persist.
    For example:
    • eBay/Amazon apps do not allow copying tracking numbers to clipboard.
    • I want to get accurate tracking information from USPS.
    • I open USPS in IE11.
    • I switch over to the eBay app and read off a few characters.
    • By the time I come back, IE11 is reloading the page, and the address bar is gone XD

    The tombstoning did not occur while apps were being switched, nor when memory needed to be cleared. The governor responsible detected that I was idly viewing material in eBay and decided to tombstone IE11.

    One of the key features of WP8.1 was "Resuming improvements". I think most improvements are due to the fact that the OS is now less aggressive on tombstoning. This is consistent with the memory bump of the newer Lumias.

    I think the problem with tombstoning is that initially Microsoft overestimated the average developer, and the complexity of mobile apps as a whole. Just until recently MixRadio had critical bugs when reanimating! They still do, but at least they do not require me to start the app in a new state.
    04-25-2015 03:56 PM
  15. a5cent's Avatar
    "If new apps are launched after an app has been made dormant, and these applications requires more memory than is available to provide a good user experience, the operating system will begin to tombstone dormant applications to free up memory."
    That quote describes the behavior of both operating system, Android and WP. <snipped> So, there is no difference.
    When you get caught the best way out is an apology.
    I can't apologize for your selective reading of the quote, nor for you not understanding it. It doesn't say what you think it says.

    The quote specifically states that tombstoning (unloading from memory) occurs if/when new apps are launched. The docs for iOS/Android don't state that. Why? Because on iOS/Android, apps can be unloaded from memory due to any memory allocation request, which could otherwise not be fulfilled. Such requests can occur at any time during an app's lifetime, not just when it's launched. I don't know why you can't see the difference.

    You might also want to look into the concept of app memory limits. This concept is completely foreign to both Android and iOS, providing further evidence that memory management on WP is in fact very different than on iOS/Android. Understand why that concept is specific to WP, and you'll understand why you've been wrong up until now, but you'll have to figure that out on your own. We're already far too technical for what most people want to read, and I'm not about to write you a book explaining WP's memory management strategies. You'll have to do your own homework.
    Last edited by a5cent; 04-25-2015 at 07:55 PM. Reason: spelling
    Yazen and gpobernardo like this.
    04-25-2015 05:26 PM
  16. jack dempsey's Avatar
    I can't apologize for your selective reading of the quote, nor for you not understanding it. It doesn't say what you think it says.

    The quote specifically states that tombstoning (unloading from memory) occurs if/when new apps are launched. The docs for iOS/Android don't state that. Why? Because on iOS/Android, apps can be unloaded from memory due to any memory allocation request, which could otherwise not be fulfilled. Such requests can occur at any time during an app's lifetime, not just when it's launched. I don't know why you can't see the difference.

    You might also want to look into the concept of app memory limits. This concept is completely foreign to both Android and iOS, providing further evidence that memory management on WP is in fact very different than on iOS/Android. Understand why that concept is specific to WP, and you'll understand why you've been wrong up until now, but you'll have to figure that out on your own. We're already far too technical for what most people want to read, and I'm not about to write you a book explaining WP's memory management strategies. You'll have to do your own homework.
    In Windows Phone suspended apps can also be terminated when the active app requires more memory, and that doesn't cause lag or stutter. Memory operation are fast.
    For example, I have Twitter and Instagram opened. Switching between them and scrolling their timelines, I see both are loaded on memory. Then, in Instagram I go to profiles of users, and other pages, after that I switch back to Twitter, and yes, It was unloaded because Instagram needed more memory.
    In Android suspended apps are also terminated when the user launches a new app if there is not enough memory available.
    Both systems behave in the same way.

    ...
    To write a book about something you have to understand the subject, which clearly is not the case.
    04-27-2015 02:18 PM
  17. a5cent's Avatar
    In Windows Phone suspended apps can also be terminated when the active app requires more memory.
    Nope. Wrong again. WP reserves far more memory for an app than it requests on launch, precisely so that which you say happens can be avoided. That's part of what those app memory limits I mentioned are for. But I give up. Believe what you will.
    04-27-2015 04:03 PM
  18. Yazen's Avatar
    What's up with the arguments xD

    Not too complicated: There are two different types of apps, and each have different lifecycles.

    https://msdn.microsoft.com/en-us/lib.../hh464925.aspx

    Compare WPF to WinRT.

    Even if you read nothing I am sure you can intuitively determine which is more complicated just by looking at the graphs.
    04-28-2015 10:12 AM
43 12

Similar Threads

  1. Xbox 360 games and Xbox one
    By Mahmoud Hussein in forum The Xbox Lounge
    Replies: 8
    Last Post: 09-22-2015, 03:12 PM
  2. Replies: 1
    Last Post: 07-01-2015, 10:27 AM
  3. Replies: 36
    Last Post: 04-26-2015, 08:37 PM
  4. Why wifi is DEAD SLOW on lumia 535???
    By Windows Central Question in forum Ask a Question
    Replies: 3
    Last Post: 04-20-2015, 02:11 PM
  5. Lumia camera app not responding
    By Sunil kumar Teki in forum Nokia Lumia 1520
    Replies: 0
    Last Post: 04-19-2015, 05:30 AM
LINK TO POST COPIED TO CLIPBOARD