Slashdot Mirror


Firefox 4 Regains Speed Mojo With No. 2 Placing

CWmike writes "With the release of Firefox 4 Beta 7 this week, Mozilla has returned to near the top spot in browser performance rankings. According to SunSpider JavaScript benchmark suite tests run by Computerworld, the new browser is about three times faster than the current production version of Firefox in rendering JavaScript, and lags behind only Opera among the top five browser makers. Mozilla launched Firefox 4 Beta 7, a preview that includes all the features slated to make it into the final, polished version next year, on Wednesday. Beta 7 was the first to include Mozilla's new JavaScript JIT (Just In Time) compiler, dubbed 'JagerMonkey,' which shot the browser's performance into the No. 2 slot behind the alpha of Opera 11."

67 of 98 comments (clear)

  1. Ofcourse it's only at number two by Anonymous Coward · · Score: 1, Funny

    because I was first!

  2. Is news.slashdot.org borken? by Anonymous Coward · · Score: 2, Interesting

    I get 404s unless I change the url to slashdot.org/...

    1. Re:Is news.slashdot.org borken? by cheater512 · · Score: 2, Interesting

      Me too. Very strange.

    2. Re:Is news.slashdot.org borken? by tehcyder · · Score: 1

      While there’s supposedly no such thing as a stupid question, that definitely comes close.

      Huh?

      --
      To have a right to do a thing is not at all the same as to be right in doing it
    3. Re:Is news.slashdot.org borken? by Crudely_Indecent · · Score: 1

      I had the same issue for a while this morning. I could get to every other /. article except this one.

      --


      "Lame" - Galaxar
  3. Realistic tests? by PhrostyMcByte · · Score: 1

    Isn't there some test that's like a consistent snapshot of popular websites, with some automation to navigate around them like a normal person would? You know, a performance test that actually might matter?

    1. Re:Realistic tests? by Anonymous Coward · · Score: 2, Interesting

      No, there's not, because the outcome would inevitably be "every browser is fast (and you won't notice a difference) given enough RAM, CPU power etc". In reality, it's not things like Javascript microbenchmark performance that matter: browsers CAN be faster or slower, but whether you can get a few ms less on Sunspider is going to be lost in the noise. The real differences will stem from other factors.

      As such, the whole thing is basically a giant red herring. I'm not sure why Mozilla is playing along here, either, although I guess they didn't manage to get the word out that these tests are a load of bollocks for real world browsing.

    2. Re:Realistic tests? by InvisiBill · · Score: 2, Insightful

      For standard website browsing, the biggest bottleneck is probably going to be the network connection. If you're talking about a handful of seconds to download a page, then a few milliseconds here and there on some Javascript isn't going to matter, like AC said. However, AJAXy webapps that make extensive use of Javascript may show a lot more difference if each Javascript function takes an extra 100ms to run. I think that's the point of these JS benchmarks - everything else is pretty much the same, so we're testing the one thing that could make a difference on script-intensive sites. Whether or not we actually have sites right now that are advanced enough to really show a difference, I can't say.

    3. Re:Realistic tests? by Anonymous Coward · · Score: 4, Insightful

      No, there's not, because the outcome would inevitably be "every browser is fast (and you won't notice a difference) given enough RAM, CPU power etc". In reality, it's not things like Javascript microbenchmark performance that matter: browsers CAN be faster or slower, but whether you can get a few ms less on Sunspider is going to be lost in the noise. The real differences will stem from other factors.

      That's because you're looking at it wrong. JS performance isn't going to improve the current browsing experience, that is perfectly true. The part you are missing is that the JS performance matters for FUTURE browsing.

      Why, you ask? One word: Flash. HTML5 exists purely to replicate most of Flash in native HTML. Flash is heavily scriptable and runs complex apps (see any heavily produced Flash game) with passable performance (most of the performance problems with Flash are due to the way the graphics stack was designed [it sucks]). If Canvas, Video and Audio tags are going to replace Flash, the JS needs to be comparable to Flash's ActionScript performance — preferably better to insentivize migration.

    4. Re:Realistic tests? by CastrTroy · · Score: 1

      I don't know. Personally, I find that IE (6,7, and 8) on my XP machines to be terribly slow. Even simple things such as opening a new tab cause quite a bit of lag. I'm not sure how it keeps up in raw JS execution speed, but the experience as a whole is pretty slow. Firefox, Chrome, Safari, and Opera on the same machine are all sufficiently fast, except for Firefox's startup time. As are the more recent versions of IE on Vista/7 are also sufficient. I do understand what you are saying, that most browsers are good enough, but there are still some slow ones out there that give a bad experience. And even though Firefox 3 is fast enough, I have to say that using Chrome, just feels an extra level quicker, and I'll use it whenever I can.

      --

      Anthropic principle: We see the universe the way it is because if it were different we would not be here to see it.
    5. Re:Realistic tests? by tehcyder · · Score: 1

      How about if you don't give a flying fuck about Flash anyway? Not (just) because it's an evil closed-source monstrosity, but because I don't want a video-internet at all?

      --
      To have a right to do a thing is not at all the same as to be right in doing it
    6. Re:Realistic tests? by PitaBred · · Score: 1

      Shall we get off your lawn now?

    7. Re:Realistic tests? by klui · · Score: 1

      reddit loads noticeably faster using b7, esp with threads with over 500 comments.

    8. Re:Realistic tests? by drinkypoo · · Score: 1

      I recall IE6 as being pretty speedy at the time, but not working right if you followed standards of course. These days anything is faster than IE on XP, though.

      --
      "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
  4. This was coming for a while... by thijsh · · Score: 1

    If you look at the graph you see this was coming for a long while now, the x64 version will also be fastest in a short while (presumably second to Opera - I don't know their x64 speed - but on windows Firefox x64 would be #1).

    This is a great achievement from the development team who consistently improved the performance from 'quite bad' to 'competetively fast', so kudos to the developers. Please don't stop now, it seems you can still stretch performance quite a bit...

    1. Re:This was coming for a while... by icebraining · · Score: 1

      Adding the JIT was obviously improve performance massively. Now the path for JS optimization is much more difficult.

    2. Re:This was coming for a while... by Rockoon · · Score: 1

      Opera doesnt have a 64-bit binary.. at least for Windows.

      --
      "His name was James Damore."
    3. Re:This was coming for a while... by thijsh · · Score: 1

      Hence my conclusion that on *Windows* Firefox x64 would be #1. ;)

    4. Re:This was coming for a while... by Rockoon · · Score: 1

      Well, I imagine that most Javascript code ends up being pointer-heavy when compiled, and I'm not sure that the extra registers are going to offset the downsides of double-sized pointers.

      --
      "His name was James Damore."
  5. Re:In 3, 2, 1 ... by icebraining · · Score: 1

    Which will be stupid, since the second page of TA is all about hardware rendering, which works even (partly) on XP, unlike IE9, and on the MacOSX.

  6. Verified with my SW-only Javascript 3D renderer by ttsiod · · Score: 3, Interesting
    Verified with my SW-only Javascript 3D renderer, try it on your own here, or just look at the submitted benchark results:

    On my aging PentiumD/2.8GHz:

    • 13 frames per second with Firefox 3.6
    • 18 frames per second with Chrome 7
    • 27 frames per second with Opera 10.6
    • 44 frames per second with Firefox 4.0beta
    1. Re:Verified with my SW-only Javascript 3D renderer by __aardcx5948 · · Score: 1

      What about the latest Chrome? 9.x

    2. Re:Verified with my SW-only Javascript 3D renderer by ttsiod · · Score: 1
      I am a newbie, Javascript-wise - so if this is true, it was purely coincidental :-)

      It's just plain math and a "plotpixel" function... so I'd be happy to know what to change to make it run faster in other browsers - any suggestions?

      (the good thing about Javascript - the code is open-source by nature)

    3. Re:Verified with my SW-only Javascript 3D renderer by ttsiod · · Score: 1

      Run Firefox4 and Chrome9 on your machine, and report back... :-)

    4. Re:Verified with my SW-only Javascript 3D renderer by rsborg · · Score: 1

      Chrome 9 is dev. The analagous test would be with Chrome 8 (beta). On my old-ass "IBM" thinkpad T60 (centrino duo w/ ATI gfx), I get 20fps on Chrome8, and 24fps on FF 3.6.12.

      --
      Make sure everyone's vote counts: Verified Voting
    5. Re:Verified with my SW-only Javascript 3D renderer by PitaBred · · Score: 1

      I cracked your top 15 with 67.84fps with 64bit Firefox from 10/15 (it's the only one stable with Flash still) and a mobile i5-520M. Just FYI ;)

    6. Re:Verified with my SW-only Javascript 3D renderer by hkmwbz · · Score: 1

      That's probably due to hardware acceleration. But you really should test Opera 11 and Chrome 8.

      --
      Clever signature text goes here.
  7. URL doesn't work by __aardcx5948 · · Score: 2, Interesting

    The link on the front page goes to http://news.slashdot.org/story/10/11/12/037241/Firefox-4-Regains-Speed-Mojo-With-No-2-Placing which doesn't work. The same URL without "news." works.

    1. Re:URL doesn't work by sgbett · · Score: 2, Funny

      216.34.181.48 news.slashdot.org

      in your hosts file also works

      --
      Invaders must die
    2. Re:URL doesn't work by metrix007 · · Score: 1

      Why would you do that, when simply backspacing part of the url is so much simpler?

      --
      If you ignore ACs because they are anonymous - you're an idiot.
    3. Re:URL doesn't work by Haedrian · · Score: 4, Insightful

      Because we're nerds, and we always look for the overly complicated solution to every problem.

    4. Re:URL doesn't work by clone53421 · · Score: 1

      Naturally, because an ugly kludgy hack due to a temporary problem is always preferable to something simple and effective.

      Along that line, is there a GreaseMonkey script to fix idle.slashdot.org? Backspacing over “idle.” is too easy.

      --
      Alexander Peter Kristopeit bought his basement from his mommy for one dollar.
    5. Re:URL doesn't work by sgbett · · Score: 1

      I very much agree with this statement. However in this specific case editing /etc/hosts barely qualifies as overly complicated!

      --
      Invaders must die
    6. Re:URL doesn't work by sjorford · · Score: 1

      127.0.0.1 idle.slashdot.org

      in your hosts file also works

    7. Re:URL doesn't work by clone53421 · · Score: 1

      If I just didn’t want to go there, I’d just adblock all links pointing me to it like I adblocked (#a(href*=goatse.)) links taking me to goatse.cx.

      --
      Alexander Peter Kristopeit bought his basement from his mommy for one dollar.
    8. Re:URL doesn't work by Crudely_Indecent · · Score: 1

      Alt+F4 speeds it up considerably, unless it crashes your browser. If you crash, try it again until it works.

      --


      "Lame" - Galaxar
    9. Re:URL doesn't work by clone53421 · · Score: 1

      That doesn’t work at all on this webpage, though.

      (Don’t use IE. No, it’s not goatse.)

      --
      Alexander Peter Kristopeit bought his basement from his mommy for one dollar.
    10. Re:URL doesn't work by Crudely_Indecent · · Score: 1

      That would have been moderately cool if it had worked.

      --


      "Lame" - Galaxar
    11. Re:URL doesn't work by clone53421 · · Score: 1

      It works on Firefox.

      --
      Alexander Peter Kristopeit bought his basement from his mommy for one dollar.
    12. Re:URL doesn't work by clone53421 · · Score: 1

      With a few caveats, I suppose I should add.

      The page has to have focus. If the cursor is in the address bar, all hotkeys work as advertised.

      Also, for some weird reason, while alt-F4 and ctrl-W don’t do anything, ctrl-F4 still works (it closes the tab, not the window).

      --
      Alexander Peter Kristopeit bought his basement from his mommy for one dollar.
  8. I love such stories by mseeger · · Score: 1

    I love stories who run like this: "Not-released-software A is faster than Not-released-software B". Great news! Carry on!

    1. Re:I love such stories by tepples · · Score: 1

      Someone who posted a comment to a previous Slashdot story about JS performance explained it to me like this: By the time you've finished developing whatever web application you're working on, the browsers mentioned in the article will have been released (barring DNF style incompetence). Deployment on end user machines is another matter, especially across major version changes.

    2. Re:I love such stories by FooBarWidget · · Score: 1

      Except the code is right out there for you to obtain.

  9. Test it yourself and see by samfisher5986 · · Score: 1

    Grab an average computer, browse with Firefox, then browse with Opera or Chrome. You'll quickly see that these benchmarks mean nothing unless you are visiting a VERY javascript heavy website.

  10. Firefox 4 beta 7 uses twice memory as 3.6.12 by microbee · · Score: 1

    Plus the Aros interface of the menu/tab bars really makes the fonts unreadable

    1. Re:Firefox 4 beta 7 uses twice memory as 3.6.12 by Velorium · · Score: 1

      +2 on the memory thing. I'm experiencing it on both Win7 and OS X with all addons disabled. I'm starting to consider other browsers, most likely Opera.

  11. Re:In 3, 2, 1 ... by Waffle+Iron · · Score: 1

    Cue the "Performance in browsers isn't just about Javascript!" comments.

    Here's an example: I switched from Firefox to Chromium last year, but not because of Javascript performance. I did it because Firefox had a nasty habit of doing a bunch of housecleaning when it shut down (I don't know, but I assume it was doing huge updates on an internal SQL database or something like that.) That could cause it to lock up for many seconds after I hit the close button; sometimes the window manager would pop up a dialog asking me if I wanted to terminate the unresponsive app.

    That behavior was infuriating, and Chromium always closes immediately. It was nice that Chromium has snappier page rendering, but it's not such a huge difference that it's totally compelling. If Firefox has addressed the shutdown issue in recent versions, I'd consider going back because the Chromium UI is so "stripped down" that they lack a few handy features that Firefox had.

  12. Re:In 3, 2, 1 ... by Anonymous Coward · · Score: 1, Interesting

    I stopped using Firefox because it would use up large amounts of memory after only a couple of days and then start lagging every 30 seconds or so until I restarted it. That is related to page rendering performance since the lag directly affected it.

  13. Javascript speed is nice and all... by Cochonou · · Score: 1

    But when it comes to Firefox, I'd rather have a plugin that would allow the use of the operating system codecs (if available) to play HTML5 video encoded in h264. That, along with the built-in support for Ogg Theora and WebM, would allow us not to care about codecs from a user perspective.

    1. Re:Javascript speed is nice and all... by bunratty · · Score: 1

      http://userscripts.org/scripts/show/70028 allows Firefox to use the Flash plugin to play HTML5 H.264 video.

      --
      What a fool believes, he sees, no wise man has the power to reason away.
    2. Re:Javascript speed is nice and all... by lightversusdark · · Score: 1

      Mod parent ironic.

      --
      "There is nothing nice about Steve Jobs and nothing evil about Bill Gates." - Chuck Peddle
  14. why do i care about speed when by cinnamon+colbert · · Score: 1, Insightful

    the organizer/editor for bookmarks is so bad (eg, where is the export a folder of book marks to email function, where is the scan for rendundant book marks button, where is..)
    when nevercooky is not in the default install
    when the new addon webpage looks like a commercial for useless crap, instead of a guide to all of hte addons
    when pdf handling still sucks in the default,and is, in my hands, number one cause of crashes
    the browser lets other people see stuff like what type of browser I am running, and doesn't try to obscure this
    new versions break old addons . . .

  15. Minor nitpick by Inf0phreak · · Score: 1

    It's "JaegerMonkey" or "JägerMonkey" if your keyboard has umlaut available. But certainly not "JagerMonkey" - q.v. https://wiki.mozilla.org/JaegerMonkey

    --
    ________
    Entranced by anime since late summer 2001 and loving it ^_^
  16. Re:In 3, 2, 1 ... by tehcyder · · Score: 2, Insightful

    But performance in browsers isn't just about Javascript!

    --
    To have a right to do a thing is not at all the same as to be right in doing it
  17. I've Been looking Forward to This by BJ_Covert_Action · · Score: 1

    I hope this performance improvement really shows in the release version of the program, and I hope it runs fairly lightweight on system resources. I've been using Firefox since my high school days and I really loved it up until a year ago or so. About the time Chrome came out, Firefox was eating up so many of my system resources (so what if I use six year old hardware, it's adequate!) that it was making it hard to get things done on my computer and have Firefox running simultaneously. I switched to Chrome for it's speed and small footprint but, honestly, internet with scripts enabled and crappy half-assed ad-blocking software is just too annoying. I'd love to return to Firefox soon. So here is crossing my fingers and hoping that the browser will run on an old Athlon 3200 processor with a single gig of pre-DDR2 memory.

    1. Re:I've Been looking Forward to This by D+Ninja · · Score: 1

      The problem is, the performance improvement to the Javascript engine does not equate to running light on the system resource usage. My guess is you aren't going to see all that much of a difference there. (However, as an avid Chrome user, I haven't actually verified my claims with Firefox.)

  18. Re:In 3, 2, 1 ... by obarthelemy · · Score: 1

    Cue "we don't really care about performance", rather. I care about feature, compatibility, safety, stability... and not a jot about performance.

    --
    The Cloud - because you don't care if your apps and data are up in the air.
  19. How about a UI benchmark? by Anonymous Coward · · Score: 1, Interesting

    Sure JS speed is nice, but the reason I switched to Chrome was simple, the UI is actually responsive.

    Launching new windows/tabs in FireFox is noticeably slow, whereas Chrome is almost instant. Rearranging tabs, slow on FireFox, snappy on Chrome.

    Try loading a large page in FireFox, while its being rendered the entire UI locks up tight. When I do Google searches I tend to open 10-20 tabs at once with the results I find interesting, then go back and look at them all. Again with FireFox this is a slow operation with the UI locking up everytime a new tab opens, whereas with Chrome it feels natural with no slowdown at all.

  20. Who cares? by ConceptJunkie · · Score: 1

    You know what? If a page takes 200 milliseconds longer to run because JS performance isn't quite as fast who cares?

    None of these browser speed wars addresses the problem that when I watch Hulu on machines that aren't top-of-the-line state-of-the-art, the video is jumpy often to the point of unwatchability.

    My pipe is fat enough and the computers I'm using can do fullscreen video just fine. It's Flash, especially on Linux, that kills performance. Most video sites still use Flash, as do a lot of those fun little games, etc., and the only only alternative to Flash in most cases is to go without the site or functionality. Flash is what makes web browsing slow. Every other performance issue is like line noise in comparison.

    --
    You are in a maze of twisty little passages, all alike.
  21. Re:In 3, 2, 1 ... by BZ · · Score: 1

    People _say_ that, but when they vote with their feet (or mice, as the case may be) it turns out that they do care about it. Often subconsciously.

  22. Re:In 3, 2, 1 ... by Tharsman · · Score: 1

    I keep a personal compilation of benchmark results from Peacekeeper's browser benchmark. I do not like their averaging of tests but I do like their individual tests. Due to this I keep the numbers in a Google Docs spreadsheet where I can easily look at them.

    Here are my results, with Firefox 4b7 included. Beta 7 does not seem to be much faster than Beta 6, although in some tests it is a big leap from Firefox 3.6.8.

    BTW, these are all ran on the same machine with no configuration changes and the same background processes running for all tests.

    I'll see if i run SunSpider's benchmark later and add it to the versions I still have on my machine.

  23. Re:Good job kids by metrix007 · · Score: 1

    Firefox has a better extensions framework, better configurability, a better interface and better performance.

    --
    If you ignore ACs because they are anonymous - you're an idiot.
  24. How about usability by rhomp2002 · · Score: 1

    I have had faster browsing times with Opera as well. My problem with Opera is that is not as usable as FF or Chrome. My biggest bugaboo with Opera is in trying to copy and paste from the window to my email so that I can send info to a friend. I do a lot of that and I cannot do it with Opera. It just will not copy and paste outside the browser itself. It works well from screen to screen within the browser but it just does not work from browser to message or anywhere else outside the browser itself. That makes it practically useless to me. Chrome and FF both do this very well. So long as I have relatively good speed and can copy and paste I am good to go, otherwise forget it. Used to have a problem with a lot of websites would not display correctly on Opera, especially if they were designed for IE but Opera seems to have solved that one in the latest release. This came up with real estate sites mostly that would not display the selection screen. Now if they just solve the copy and paste then I will bring it up as a potential for default browser.

  25. Evolution of JavaScript Implementation Speed by RAMMS+EIN · · Score: 1

    This story makes me curious: how has JavaScript implementation speed improved over time? I see a lot of benchmarks comparing recent versions of browsers, but does anyone have a comparison against, say, Firefox 1.0? Also, how do current JavaScript implementations stack up against current implementations of other languages, such as C, Lua, or Python?

    --
    Please correct me if I got my facts wrong.
  26. At cost of? by Ilgaz · · Score: 2, Insightful

    In other news, OS X PowerPC was dropped and there is also a crazy talk about dropping anything=SSE3. They really lost the design philosophy of Firefox or anything Mozilla. What happened to portability especially when nobody likes their mobile offerings? How many times they must learn to think outside X86 PC?
    Firefox in speed race with a company who controls/knows about every single device they shipped (Apple) and another who can spend couple of billions of dollars in no time without even noticing it (Google).
    As a person who knows and always respected Mozilla mission, I feel obliged to post this although I know what kind of karma suicide it is.
    I really wonder what kind of management took over Mozilla. I remember old days these guys spending hours to make it run/compile on OS/2, dead desktop OS, not completely documented and stabilized one like OS X 10.5.8 PPC.
    What about Atom CPUs Mozilla guys? Drop that too as it will make you score less in couple of benchmarks? I heard netbooks are outfashioned, perhaps you should drop them next to keep yourselves cool (!)

  27. I respect Adobe 100x more than them by Ilgaz · · Score: 1

    You know what? If a page takes 200 milliseconds longer to run because JS performance isn't quite as fast who cares?

    None of these browser speed wars addresses the problem that when I watch Hulu on machines that aren't top-of-the-line state-of-the-art, the video is jumpy often to the point of unwatchability.

    My pipe is fat enough and the computers I'm using can do fullscreen video just fine. It's Flash, especially on Linux, that kills performance. Most video sites still use Flash, as do a lot of those fun little games, etc., and the only only alternative to Flash in most cases is to go without the site or functionality. Flash is what makes web browsing slow. Every other performance issue is like line noise in comparison.

    Old times, when some company abandons/will abandon your computer/OS or you can't decide whether to join the herd (windows) or run linux/BSD, you always had Firefox at your mind. You would think it would have Firefox support and would run it one way or another.
    Now, they dropped PowerPC binaries (because their cool looking addressbar not working) and speak about dropping anything below SSE3. Publicly that is...
    Code is being infested by completely unportable x86 specific ASM to join cool kids with JS asm acceleration.
    If they weren't afraid of Win32 users, they would drop X86 32bit in no time because 64bit runs 20% faster!" and they will have great fan feedback from idiots who have no clue about anything other than 64bit is faster.
    Does Adobe support PowerPC? Yes they do. They have to do massive trickery on OS X (not hacks) but they do. Your Flash is probably slow because it doesn't do GPU decoding. If you use Linux, Adobe has a weird excuse. If you use OS X, they have a very good/justified excuse as Apple refused to add a central GPU decoding framework open to others (not private) until OS X 10.6.3. Running 10.5.8 on a production machine you can't update? Bad luck.

  28. Re:In 3, 2, 1 ... by ElderKorean · · Score: 1

    But performance in browsers isn't just about Javascript!

    Whichever browser can run Farmville/Fronterville/Mafiawars (all facebook apps) the best will win over most people.

    I have a modern-ish machine that runs Lord of the Ring Online quite well, but crawls when I looked up those apps.