HTML5 performance in Apple's iOS 5 tops Windows Phone 'Mango'

13»

Comments

  • Reply 41 of 51
    macrulezmacrulez Posts: 2,455member
    deleted
     0Likes 0Dislikes 0Informatives
  • Reply 42 of 51
    Quote:
    Originally Posted by MacRulez View Post


    Have you never seen any Flash game run at 60FPS?



    Yes, the Google team's efforts at porting QII are laudable, apparently even here is this vehemently anti-Google forum.



    But unless you can find an equivalent build by developers of similar expertise in Flash, it's a fine showpiece but not a comparison.



    Like I said, I don't use Flash and have far less invested in it than the company hosting this forum, and I agree that HTML5 is coming along nicely.



    But if you think it's fast now, you can't imagine how much faster it could be if it used bytecode instead of having to parse JS and XML on the fly.



    Equivalent in Flash? There's enough online multiplayer games on the internet - pick one and go for it.
     0Likes 0Dislikes 0Informatives
  • Reply 43 of 51
    Quote:
    Originally Posted by MacRulez View Post


    If you can honestly say that your battery runs as long playing QII in your browser as it does showing static web pages, you will have found your magic faery and I will happily acknowledge being corrected.



    What? You are going completely off topic here. You said find an equivalent to Flash in interactivity and graphics, so I showed you Quake II. Then you said show battery life, so I showed activity monitor shots with brief explanations.
     0Likes 0Dislikes 0Informatives
  • Reply 44 of 51
    macrulezmacrulez Posts: 2,455member
    deleted
     0Likes 0Dislikes 0Informatives
  • Reply 45 of 51
    Quote:
    Originally Posted by MacRulez View Post


    Ah, thanks for clarifying that the XML comment was just a straw man, since you acknowledge that, in contrast to HTML5, Flash Actionscript is compiled to more compact bytecode and that only the subset of Flash SWFs that may need additonal AJAX requests would be parsing XML.





    Preaching to the choir, my friend. Unlike yourself and Apple Insider, I've never been dependent on Flash.



    As I mentioned in the very post you're quoting, where one might presume you read this before hitting Reply:





    If you can show me complex animations in HTML5 that are as kind to your battery life as static pages, you will begin to have a relevant point.



    In the meantime, I fully support HTML5 and have no stake in Flash. I merely acknowledge that any system that can move computation from runtime to compile-time will have an inherent advantage. Whether Adobe is in a position to make full use of this inherent advantage against the tens of millions being poured into ECMAScript by Google, Apple, Microsoft, and others remains to be seen, and as an HTML5 fan it doesn't concern me.



    The bigger concerns are about perceptions of the impact of complex renderings on battery life.



    As Flash use continues to wane, loyal readers here will wake up one day to a world in which superfluous animations and annoying ads are made in HTML5, and they will see the hit to their battery life.



    But unlike Flash, there will be no way to turn if off without turning off JavaScript itself, rending all web apps useless.



    Pick your poison...



    Stop going off topic. I never said I was dependant on Flash, I never said you didn't care about HTML5.



    Oh look, google found a great animation example in HTML5. http://9elements.com/io/projects/html5/canvas/

    Comparable to Flash? Oh my yes! Its running perfectly fine on my Celeron Based Windows 7 Laptop in full screen (1366x768) in IE9 at around 60% CPU usage. A 720p (slightly lower resolution) youtube video (far less demanding task) in flash is using 55% of the CPU. Wow... a video, using THAT music CPU power? Yet a fully interactive HTML5 demo, graphics rendered on the Fly, with music, better than 720p, is only 5% more CPU power than Flash playing back a 1:17 long video file?



    I'm saying no more on the subject, I've given evidence and yet you've supplied none, so I'll just stop waisting my breath.
     0Likes 0Dislikes 0Informatives
  • Reply 46 of 51
    macrulezmacrulez Posts: 2,455member
    deleted
     0Likes 0Dislikes 0Informatives
  • Reply 47 of 51
    firefly7475firefly7475 Posts: 1,502member
    Quote:
    Originally Posted by benanderson89 View Post


    I've given evidence and yet you've supplied none.



    You need to run an apples to apples comparison. One exists here.
     0Likes 0Dislikes 0Informatives
  • Reply 48 of 51
    Quote:
    Originally Posted by Firefly7475 View Post


    You need to run an apples to apples comparison. One exists here.



    I fired up my windows laptop (a Celeron T3500 system) and gave it a whirl in fullscreen (1366x768) in IE9 (so HTML5 and Flash were hardware accelerated). The Frame rate claims of the flash test are bogus, you can clearly see with the human eye that its the same or less than the HTML5 test - its dropping frames and stuttering like no tomorrow.



    That said:

    Canvas Test - 50% CPU Usage, spikes to 60% at times.

    Flash Test - 50% CPU Usage, spikes to 80% at times, 200MB More RAM than Canvas test in use.

    HTML Test - did not render correctly in IE9.

    SVG Test - See canvas test
     0Likes 0Dislikes 0Informatives
  • Reply 49 of 51
    firefly7475firefly7475 Posts: 1,502member
    Quote:
    Originally Posted by benanderson89 View Post


    I fired up my windows laptop (a Celeron T3500 system) and gave it a whirl in fullscreen (1366x768) in IE9 (so HTML5 and Flash were hardware accelerated). The Frame rate claims of the flash test are bogus, you can clearly see with the human eye that its the same or less than the HTML5 test - its dropping frames and stuttering like no tomorrow.



    That said:

    Canvas Test - 50% CPU Usage, spikes to 60% at times.

    Flash Test - 50% CPU Usage, spikes to 80% at times, 200MB More RAM than Canvas test in use.

    HTML Test - did not render correctly in IE9.

    SVG Test - See canvas test



    I get similar results, but with less CPU usage overall and not so many spikes with Flash.



    I think the IE9 canvas render looks smoother than Flash as well. I haven't done a proper study though.



    He also has a benchmark that limts to 25fps here. It's probably an even fairer test for CPU/memory usage.
     0Likes 0Dislikes 0Informatives
  • Reply 50 of 51
    Since this story broke, it has widely been reported that the WP7 test was from March. More recent verisions of Mango have consistently outperformed iOS 5 http://wmpoweruser.com/ios5-faster-t...ndering-wrong/

    with 41 FPS.
     0Likes 0Dislikes 0Informatives
  • Reply 51 of 51
    solipsismsolipsism Posts: 25,726member
    ACHTUNG: Do not click the link in the above post.
     0Likes 0Dislikes 0Informatives
Sign In or Register to comment.