The Latest Web Browser Grand Prix

An anonymous reader writes "The latest browser benchmarks are in... again. This is one of the better 'browser battle' articles, though. Chrome 13, Firefox 6, IE9, Opera 11.50, and Safari 5.1 are put through 40-some tests on both Windows 7 and Mac OS X Lion. As a PC guy, I was pretty impressed with the performance of Safari on OS X, and the reader feature looks awesome too. The author also uncovered a nasty Catalyst bug that makes IE9 render pages improperly and freeze up under heavy loads of tabs. The tables at the end pinpoint the strengths and weaknesses of each browser, which is nicer than a 1-10 or star rating. The tests are more thorough than most browser comparisons I've seen."

Comments

Noscript?

[Anonymous Coward]

There is only one important question; Does it run Noscript?

Re:Noscript?

[beelsebob]

Sure, I'm not sure I'm aware of a modern browser other than IE that can't run a javascript blocker.

Re:Noscript?

[Anonymous Coward]

except that noscript in chrome will make chrome actually run slower... not to mention heavier addons like adblock

Re:Noscript?

[m1ndcrash]

Native Block Content in Opera

Re:Noscript?

[DaVince21]

I answer with a more important question: does it need to run Noscript?

Chrome for example can block individual scripts all by itself. No extension required. Of course, something to handle clickjacking and those kinds of tricks would still be nice.

Why does this matter?

[Anonymous Coward]

Okay, we're no longer in the days where your 486SX with 2MB of RAM would suffer greatly if your code was interpreted versus compiled in hand-crafted ASM.

For the average user who has more RAM and CPU cycles than what they know what do with; does it matter that a page will load in 1sec or 2sec? Seriously, 99% of the consumers won't care.

Look at the computer from the next casual person you have? You'll notice that they're using 5% of their RAM and 2% of their cpu(s).

Efficiency matters in an enterprise environment or for power users (gamers, video/audio, new ways to create shell sorts).

Re:Why does this matter?

[Lisandro]

Look at the computer from the next casual person you have? You'll notice that they're using 5% of their RAM and 2% of their cpu(s).

If only. Try firing up Firefox with 10-12 tabs and see it slowly, but steadly, eating you memory up. A browser is one of the many apps i run on my systems, so good peformance and memory handling has a definite impact on my user experience.

Re:Why does this matter?

[jhoegl]

Agreed
The only things I care about

General:
Stable

Personal:
Layout
Plugins
Applies standards correctly

Professional:
Restrictions
GPO availability

Re:Why does this matter?

[royallthefourth]

Try firing up Firefox with 10-12 tabs and see it slowly, but steadly, eating you memory up.

Good! RAM is made to be used. This means your computer is working as intended.
Stop watching system graphs and just use the computer.

Re:Why does this matter?

[Lisandro]

It's hard to use the computer when you find up the damn browser is eating half of your 4 Gigs of RAM :) I like Firefox overall, but they really need to start addressing their memory management issues.

Re:Why does this matter?

[Anonymous Coward]

Upgrade to 8gb.

With two browsers (firefox and chrome), pidgin, teamspeak, world of warcraft, msi afterburner, and a dozen other apps running, I've never seen my machine top about 6gb of ram usage.

Re:Why does this matter?

[Anonymous Coward]

How quaint. A spec nerd. Hint: nobody cares about your cores or your trigahertz.

Internet Explorer comes preinstalled on over 90 percent of computers sold yet only about 50 percent of computers on the internet are running it. Therefore, a signicantly more than 1 percent of the population has made a determined choice to use something else. That blows your 99 percent stat out of the water right there, bucko.

Look at the computer from the next casual person you have? You'll notice that they're using 5% of their RAM and 2% of their cpu(s).

The same could be said for cars. Most cars will go from 0-60 in under 10 seconds and will do over 100 miles per hour. How often do you need to do either of those things? Seldom. But when you do, you do. When people need their processor to peg, say for a javascript heavy site like yahoo.com, google maps, facebook, etc, they will appreciate a faster browser.

Efficiency matters in an enterprise environment or for power users (gamers, video/audio, new ways to create shell sorts).

Fucking elitist prick. Shut your hole.

I think you forgot which site this is. read the site motto. "News for nerds". Please go back to digg or reddit.

Re:Why does this matter?

[DrXym]

If only. Try firing up Firefox with 10-12 tabs and see it slowly, but steadly, eating you memory up. A browser is one of the many apps i run on my systems, so good peformance and memory handling has a definite impact on my user experience.

Fire up 10-12 tabs and chances are you have multiple instances of Flash bogging down your computer. If not flash then you still have 10-12 DOMs, 10-12 JS sessions with random timer events, image animations and so forth going on in the background. I think Firefox should probably ship with something analogous to a Task Manager where you could see how much CPU each tab "consumed".

I don't even blame Flash for the problems of CPU consumption. Flash gets a lot of hate but its really a victim of its own success. Any piece of code which had so many instances running would hog CPU.

Re:Why does this matter?

[Anonymous Coward]

How quaint. A spec nerd. Hint: nobody cares about your cores or your trigahertz.

You are a moron. You just proved the guy correct. That's the whole point. A browser just ain't cpu-intensive. Doesn't matter if you got a 6-year-old rig or the latest and greatest.

Internet Explorer comes preinstalled on over 90 percent of computers sold yet only about 50 percent of computers on the internet are running it. Therefore, a signicantly more than 1 percent of the population has made a determined choice to use something else. That blows your 99 percent stat out of the water right there, bucko.

You're silly. And how many installed it because it's faster? It's the UI and options that they like. Sure, it shouldn't be slow as molasses, but which browser is?

The same could be said for cars. Most cars will go from 0-60 in under 10 seconds and will do over 100 miles per hour. How often do you need to do either of those things? Seldom. But when you do, you do. When people need their processor to peg, say for a javascript heavy site like yahoo.com, google maps, facebook, etc, they will appreciate a faster browser.

A car analogy. Quick: what kind of emergency requires you to load a page super-super-quick?

Re:Why does this matter?

[Eponymous Coward]

Yes it matters. There are usability studies that show people will wait X seconds (not sure of the exact number) before they close the page and give up. If a web browser is faster, it's going to make more page views and if you are a business that makes more money when people use the web more (like Google), then it matters.

Re:Why does this matter?

[Anonymous Coward]

Wow, imagine this. A slew of "AC's" come out of the woodwork to shit on an article that concludes that Chrome beats Internet Explorer. "Oh, people don't care about speed." "People don't care about RAM." Bullshit. You astroturfing fuckers are just made that IE is a piece of shit and Chrome and Firefox are better.

Re:Why does this matter?

[Lisandro]

I usually get along using way less than that. For example, this is my (crappy) desktop at work, right now...

Re:Why does this matter?

[Sunday_Ironfoot]

For the average user who has more RAM and CPU cycles than what they know what do with; does it matter that a page will load in 1sec or 2sec? Seriously, 99% of the consumers won't care.

Amazon found that a 100ms delay decreased sales by 1%, Google also found similar results where an artificial increase in page loading time decrease the number of searches users performed. So it appears that users DO care, at least at an unconscious level.

Also, you're partly missing the point of faster browsers. As browsers get faster & more advanced, web developers will find interesting ways to take advantage of that extra power and capabilities and deliver more compelling user experiences (look at 3D on the web with WebGL, or ultra low latency two way communication with WebSockets for example). Unless you're one of those Web Luddites who thinks the web should just be black text on the white background with the occasional image.

Re:Why does this matter?

[Lisandro]

I mentioned Firefox specifically because i can open a crapload of tabs on other browsers without these issues. I've posted a pic of my workstation upper in the thread, where Opera is shown behaving very nicely in these situations.

Again, i like FF a lot. The developers seem have started addressing these issues since version 6, but still, i keep finding out it leaks memory like crazy after a while.

Re:Why does this matter?

[Anonymous Coward]

You need to take your medication, son. You're going to pop a vessel, my fellow AC. Or at least go hang out at different web site.

I use Firefox if that makes you feel better. On Ubuntu. I still don't care if a browser is 20% faster and that it uses more of my expensive gold-plated RAM. You're still being silly though.

Re:Why does this matter?

[oakgrove]

What about the people that are going to be running ARM tablets? The browsers on those use the same rendering engines as the desktop counterparts, i.e., Webkit, Gecko, and Trident. And the javascript engines are the same as well. We need those engines to be as efficient as possible to eek out better and better battery life. Obviously, this benefits smartphones as well. The entire world doesn't revolve around the desktop.

Re:Why does this matter?

[oakgrove]

You don't really think facts and rational thought are welcome here do you? This is Slashdot, we'd much rather go for the easy populist +1 by criticising every little thing whether it actually improves our lives. After all, what you propose amount to change. And we can't have that now can we?

Re:Why does this matter?

[Anonymous Coward]

Fire up 10-12 tabs and chances are you have multiple instances of Flash bogging down your computer

Nope.

NoScript is good. I prefer PrefBar. Single-click browser-wide (and in some cases, per-tab) activation/deactivation of images, cookies, Java, Javashit and Flash. Idiot webmaster does browser-sniffing based on User-Agent? Forge it with a single dropdown. Don't like some asshat web designer's choice of red text on a blue background? Heck, you can even turn off "colors".

I don't even blame Flash for the problems of CPU consumption. Flash gets a lot of hate but its really a victim of its own success. Any piece of code which had so many instances running would hog CPU.

True - but Flash probably is the memory hog here. I have an instance where I don't even have it installed (really!), and Firefox 3.6.20 has never taken more than 700MB even after weeks (!) of use with hundreds of tabs open.

The current session is about three days old, and currently uses 200MB, with 28 tabs open in four windows. Adding two image-heavy Fark photoshop threads and an 800-post hurricane thread added only 30MB to the total. Adding this extremely image-heavy 700-post Caturday thread temporarily bumped it to 400MB while rendering, but it stabilized at 300MB, and gave the RAM back when the thread was closed. Hardware is a core i7 with 6GB RAM, barely even hiccuped while rendering it. Only cheat I have is an ad-blocking proxy, and most of the time Javascript is disabled. PROTIP: If the site loads with Javascript disabled, enabling Javascript does not run the scripts on the site without a subsequent reload. That's not a bug, it's a feature: you can have dozens of tabs open to your favorite sites, and they will burn no CPU even when Javascript is temporarily enabled when you want to use something like Google Maps. Great for laptops!)

Re:Why does this matter?

[Nemyst]

So? Does your computer suffer from that? Do other programs crash or run out of memory?

Free RAM is useless, Firefox is just addressing it and leaving it addressed for future uses. If the computer needs more RAM for other applications, it just releases it THEN, instead of leaving RAM unused.

Re:Why does this matter?

[Anonymous Coward]

Is that so? Why does my computer start to crawl and become unresponsive when I've left Firefox on for a few days? Freeing that memory by closing Firefox makes it become responsive again. The only reason I ever look at graphs is to find out why my computer is no longer usable.

Re:Why does this matter?

[thetoadwarrior]

Firefox on my MBP (4 gig of ram) doesn't use half of the memory and my Ubuntu work machine has 2 gigs of ram and there is no difference in performance when I have a ton of tabs open which will include flash apps (NPR radio) and javascript intensive sites (slashdot/ reddit). If you genuinely have problems with Firefox on 4 gigs of ram then the problem probably isn't Firefox.

Re:Why does this matter?

[Lisandro]

Yeah, i guess other browsers are just magical...

Re:Why does this matter?

[Anonymous Coward]

You are an idiot. The engines in these browsers also run on cellphones and tablets. Efficiency matters where battery life is at stake. Have you ever even seen a laptop?

Re:Why does this matter?

[Anonymous Coward]

I don't think you understand how memory management works.

Re:Why does this matter?

[misexistentialist]

I wouldn't care, but Firefox becomes unusable once it hits 2GB or so.

Re:Why does this matter?

[Anonymous Coward]

Damn, you're stretching now, this must really be bothering you. You are highly entertaining and your feeble mind is no competition for many of us here.

Do you know anything about computers? You're talking about different platforms ON DIFFERENT CPUs. Your compiler will make the most difference in regards executable efficiency. Something that is great on x86 isn't necessarily going to slay on an ARM cpu.

Your touchpad app on your notebook will make a bigger difference between using IE9 and Firefox6 when it comes to power savings. Or in your case, your online chat app with your psychotherapist.

Go take a long walk on a short pier. Preferably in the eastern seaboard right now. It might cool you off and spring some sense in your head.

Re:Why does this matter?

[Anonymous Coward]

If I stop occasionally looking at the graphs and restarting the browser every so often, I tend to notice when Firefox randomly ties up the entire system in a swapping hell. It is a little late to close browser when the mouse becomes non-responsive and logging into a text terminal times out because of the excess swapping. Stuff like that means I tend to know how much memory it is using without looking at a graph, when the system is brought down by just clicking on a link, opening a blank tab, or even scrolling down on a page in a browser with flash and most javascript disabled.

Re:Why does this matter?

[Anonymous Coward]

False. Unused RAM is used for caching hard drive accesses. Even if you use an SSD, you get a performance boost from that, but if you've got a regular platter drive? The performance you lose when RAM is being gobbled up instead of used for HD caching just because of your browser is extremely annoying.

Re:Why does this matter?

[Anonymous Coward]

your feeble mind is no competition for many of us here.

The irony is killing me.

Your compiler will make the most difference in regards executable efficiency.

No it doesn't fool. How you write your program will make the most difference in regards to executable efficiency. Then, you look at how good your compiler is. God, you are stupid. And most programs tend to have roughly equivalent relative performance when run on ARM or x86. I should know. I have an ARM device running Ubuntu 11.04 as well as a laptop/desktop/netbook, etc and slow applications on x86 are slow on ARM and fast applications on x86 are, who'da thunk it, fast on ARM. Please come back with personal experience before excreting your bullshit again. Of course, the really funny thing is you are comparing the wrong thing. We are comparing the relative performance of different code bases. I.e., the different browsers. Basically, you are arguing that code efficiency doesn't matter.

Your touchpad app on your notebook will make a bigger difference between using IE9 and Firefox6 when it comes to power savings.

Then we need efficient touchpad apps, stupid. Maybe all of our applications should keep at least one eye on efficient code bases since it all adds up. But just keep bringing the stupid. You're good at it.

As for your pathetic insults, thanks for making yourself look more like a childish brat to be ignored.

Re:Why does this matter?

[tepples]

Buy more ram

How does one buy more RAM when both slots on the motherboard are filled with the largest stick that they'll take?

Re:Why does this matter?

[epine]

No it isn't. Stop watching the graphs and you won't even know.

I once had a doctor just like you. I've had a sleep issue for twenty-five years, during which I've become more adept than your average goat at noticing certain details of my physiological state. This particular doctor implied that I was so naive about the scientific process as to verge on creationism and that I matched any wacky hypothesis to reality with no regard to observation. He told me I had no data.

Actually, what I have is an R workspace with an 80,000 line CSV file extracted from Firefox showing my browsing activity over about a year period during which the white band of "away from my desk" appearing roughly once every 24 hours migrates diagonally on the circadian plot with a one hour daily drift. I've managed to treat this subsequently with carefully timed melatonin administration and have reduced the period to roughly 24.15 hours. Miss just one sunrise control pill and I'm an hour pregnant the next morning. And since I've never had a reverse gear on this hasty blue marble, that adds up to a week of night shifts sooner or later.

The other aspect of 80,000 Firefox page requests over a one year period is that I have actually noticed Firefox being one of the worst GD memory pigs of in all of god's creation without consulting system monitor, so STFU about the system monitor. Maybe I installed too many useful extensions, but then if I didn't want the extensions, I would use Chrome instead.

With 200 tabs open in eight FF clients spread over nine desktops, after about ten days, I can often type half a dozen words during frequent FF gcgag stalls (garbage collect gag) before my text blurps out. Whenever FF virtual memory climbs to over a gigabyte, I pretty much have to close my eyes while typing, as the feedback loop in the HTML input box causes me more distress than assistance. I participated heavily in a FF beta a couple of years ago where memory usage was three times worse than it is now. I was restarting FF every few days just to clear the constipation. This on a Linux system with 4GB of memory since upgraded to 8GB.

Thanks for giving my asshole GP a nice pat on the back in his self-satisfied assessment of the observational powers of his hapless sleep-deprived clients.

Re:Why does this matter?

[arbulus]

So if one's motherboard cannot handle more RAM, are they supposed to just buy a new one? Is the fact that you have 16 GB RAM an excuse for shotty app development? "Oh, we have plenty of RAM, who cares about system resources?" I have 8 GB of RAM on my rig, but I don't want Firefox taking up 2-3 GB of it just because it can. And there's no reason it should, except for lazy developers. Your "buy more RAM" attitude is snobbish at best, and justifies lazy developers at worst.

Re:Why does this matter?

[Anonymous Coward]

you talk like a fag

So sad. You are such a hateful person.

Re:Why does this matter?

[tlhIngan]

So? Does your computer suffer from that? Do other programs crash or run out of memory?

Free RAM is useless, Firefox is just addressing it and leaving it addressed for future uses. If the computer needs more RAM for other applications, it just releases it THEN, instead of leaving RAM unused.

Yeah, I really do love how Firefox is so slow, when I alt-tab to it, it looks lik enothing happens for 2 minutes or so, then pops up.

Or it can pop up instantly, but the moment I do anything like scroll or switch tabs, it stalls and even Windows gives me the "Not responding" signal.

Free RAM is useless yes, but mis-using system resources and relying on swapfiles just makes the experience painful. An SSD helps, but until it's standard in every PC so it can mask poor application behavior, it's not desirable to waste RAM.

Use as much as you need, but don't gobble more and waste it. A small working set and enough memory to do common tasks without having to swap in 500+MB of data makes firefox far speedier than it is now. It's what makes Firefox painful to use when it gets swapped out and what people complain about when they say "don't use so much RAM".

Especially when Firefox swaps in its entire address space, you use it for a few seconds then use another app that swaps stuff in (but is otherwise responsive), forcing Firefox to swap out again and take another couple of minutes to "wake up".

Re:Why does this matter?

[Anonymous Coward]

If a given page takes more than a few seconds to load, one of three things is at fault:

1. web server is non-responsive. No browser is going to help here. site owner should focus on better hardware.

2. web page is badly designed and full of flash and ads that prevent the page from loading until they've downloaded a ton of crap. good browser extensions will make a difference here, mainly by preventing that crap from loading in the first place, which is of no help to site owner as his goal was to get you to view that crap.

3. User's computer is 12 years old, or runs like it is 12 years old due to high amount of malware, spyware, viruses, and toolbars on it. Just like case 1, no modern browser is going to be able to help here.

end result: the percentage of cases where it is the user's choice of browser that will determine whether said users shuts the page in disgust before being able to view ads is astonishingly low.

Re:Why does this matter?

[epine]

Swearing at my GP was actually helpful this time around, so I now have one positive swearing result out of N as N goes to infinity concerning this medical interaction.

I've got the data set under my thumb to show the man the back of my hand, but it's technically tricky to precisely fit negative attestation data: my FF stream tells me when I'm awake and clicking but not when my cheek is lolling on the Z key.

I just realized I can score each moment in time by linear distance to nearest click (painting my life with a snore to shore metric) and then use the R package FDA (Functional Data Analysis) to fit the data on a Fourier basis using the harmonic acceleration penalty to smooth the curve.

Harmonic acceleration makes my furrow my brow. It's defined as:
L = omega^2 D + D^3
where D is differentiation. Don't completely grasp why this works and haven't tried it out yet.

Ramsay, Hooker, and Graves mostly use it to fit precipitation data. On my personal weather channel, there would be the daily peregrinations of the sand man: the sun will be rising hours before summoned and hang in the sky much longer than usual due to an extraordinary ideation wave, followed by not very much for a day or two. Same old same old. Why do I even tune in?

Re:Why does this matter?

[Anonymous Coward]

It's more difficult to use the system when the browser consumes the memory needed by the applications you do real work with. It's more difficult to use the web browser when it periodically crashes due to address space exhaustion. This is not used memory, it is wasted memory.

Re:Why does this matter?

[Anonymous Coward]

Firefox has no way of knowing the overall system memory pressure. It cannot release memory when another application needs it.

Re:Why does this matter?

[bberens]

Page load time is incredibly important. Google, for example, has spent millions upon millions of dollars putting servers closer to users so they can knock milliseconds off the page load times. The difference is usually imperceptible to the conscious mind but does have a tendency to bring users back. Also, IMHO we're just now seeing the building blocks for full fledged 3d gaming and other "enterprise" applications running in the browser (Docs hardly counts).

Re:Why does this matter?

[Bucky24]

Wow, sounds like we've got a firefox fanboy here. I've actually never met one before. Who are you to comment on how a program is affecting someone else's computer? You have no idea of the setup (aside from the 4 GB of ram), and no idea even what operating system is installed. But you first say ram is meant to be used, as if the browser is the only program running on the system, and then simply deny the fact. I personally have seen firefox use much more than 2 GB of ram during normal usage (though normal for me tends to be more like 20 tabs). It's a horrible memory hog, and simply asserting "no it isn't go bury your head in the sand" is not going to make that change.

Re:Why does this matter?

[Bucky24]

No, the computer first sends all that "unused" ram to the paging file, THEN releases it. That takes quite a bit of time.

Re:Why does this matter?

[Eponymous Coward]

If you think the number of computers in category 3 is astonishingly low, then you are wrong. Think about what the average computer out there is? There's a decent chance it's running XP with something like 512MB of RAM. Even if it's running Vista or Windows 7, the average computer isn't going to be very good. Next consider that half of the machines out there are only as good or worse than this machine.

Re:Why does this matter?

[Anonymous Coward]

You're fucking crazy. You should see a doctor about that.

Re:Why does this matter?

[Anonymous Coward]

Except Firefox doesn't seem to be releasing memory back for other applications. At work it has this horrible habit of slowly eating all of the RAM in my system, and the more RAM it eats the slower the whole OS gets. It gets to the point that when I open other applications it takes seconds for them to actually DRAW on the screen, so every hour or two I have to restart Firefox. I would just switch to something else, but unfortunately I am not allowed to install another web browser at work, so it's either Firefox 3.x or Internet Explorer 8. I used to be a huge Firefox fanboy too, but at home I switched to Chrome a long time ago. I prefer individual tabs crashing over the whole web browser crashing.

Firefox 3.x? Yeah, well, where I work we have a very, VERY, large enterprise network (over 1 million employees)... all software has to be tested (well, is supposed to be tested) before being deployed. A very slow process. And with the new rapid release cycle I expect us to drop Firefox unless things stabilize.

(Over a million employees? Yes, http://en.wikipedia.org/wiki/United_States_Army)

Re:Why does this matter?

[chrismcb]

For the average user who has more RAM and CPU cycles than what they know what do with; does it matter that a page will load in 1sec or 2sec? Seriously, 99% of the consumers won't care.

I'm not sure what RAM and CPU have to do with it.MOST people would rather have their pages load in 1 second over two seconds.

Re:Why does this matter?

[Yamioni]

I don't know what the hell you, all my sibling posters, and my sibling's children posters are doing with Firefox. Trying to run Folding@Home from within some sort of browser shell/VM? I have Firefox running 24/7/365 with a minimum of two tabs open, and upwards of 10-12 at any given time, many with flash or video content and high memory overhead. I've yet to see the thing break 600MB of usage in the last two years. Before that, sure, it was attrocious and would nom up over 2 gigs of RAM. Maybe you all need to stop bitching about the fast dev cycle and upgrade.

Re:Why does this matter?

[Yamioni]

Amazon found that a 100ms delay decreased sales by 1%, Google also found similar results where an artificial increase in page loading time decrease the number of searches users performed. So it appears that users DO care, at least at an unconscious level.

In other news, a new study shows that eating slower on your lunch break leads to you eating less food. Story at 11.

As for the Amazon bit, the longer a person has to mull over their decision to actually buy what they added to their cart, the more likely they are to not order. Like the car salesman that tells you he has another buyer interested to get you to buy faster. The longer you think it over the less likely it is he closes the deal.

So no, the users don't care. If a person has actually made their mind up on buying something, 100ms isn't going to affect that. If anything they would favor the slower load as it might help them curb impulsive spending. You hear far more people saying 'I'm spending too much money!' than you do saying 'I can't spend my money fast enough!' Google users aren't going to give a toss if they 'run out of time' and complete fewer searches. 99% of the time the least important searches are going to be performed last, and the user isn't going sweat over not completing those searches. Anyone on a tight enough of a schedule to have all of their searches be of equal importance that must be completed or bad things happen probably isn't going to be using google in the first place. The only people helped by browsers being faster than they are now are the people trying to sell you something. It doesn't help the user at all.

Re:Why does this matter?

[Adm.Wiggin]

That's the problem: it doesn't. It eats the RAM, and keeps it eaten. If I don't "watch the graphs", then I don't know when I need to start closing tabs because it's eating too much, and my computer starts to swap and then I do notice, because at that point, everything gets painfully slow. This isn't just a problem with Firefox, however.

Re:Why does this matter?

[Anonymous Coward]

Miss just one sunrise control pill and I'm an hour pregnant the next morning.

So that's what they're calling birth control pills nowadays?

Re:Why does this matter?

[jesser]

A browser is many of the many apps i run on my systems

FTFY.

(If you're running an email app, a feed aggregation app, and a social network app in your browser, then your browser is going to use some memory.)

Re:Why does this matter?

[lpq]

Um...so if you open 50-60 pages, .... that's um....well over a minute...

That's dead slow for something like "opening 5 pages", considering, I can open 400+ mailboxes and search through them in under 4 seconds -- as
for display -- who could read that fast? But about 200K of 'moderate pages (has to skip through headers even for short messages...

So 1-2/secs/page really sucks.

and you wondered what I was gonna do w/my parallel cpu's...

Re:Why does this matter?

[gullevek]

I think you have on or two tabs too many.

Re:Why does this matter?

[Anonymous Coward]

Both Slots are full? Time to upgrade into this decade.

If ones motherboard can not handle modern applications, yes they should buy a new one. If that means they need more ram slots or a PCIEx2 Bus or whatever...

It's not snobbish it's a reality. I wanna play COD 4 on a $10 PC I picked up on EBay, or watch blue rays, unfortunatly I can't - must be cause of lay-z deveopers?

Re:Why does this matter?

[DaVince21]

Yes. Yes, it matters. I have a dual core 1.2Ghz laptop and some HTML5 (and actually Flash too) games still run like shit. They're just simple game, I know they have to perform better.

Oh, and page loading isn't too great either. Especially when I have many tabs opened.

Re:Why does this matter?

[DaVince21]

Addressing their memory? :P

So who won?

[Joce640k]

Am I going to have to read the article to find out who won?

Actually, it doesn't matter: Only one of them runs adblock and noscript...'nuff said.

Re:So who won?

[Anonymous Coward]

Chrome and Firefox both use Adblock, for the record. Chrome doesn't have "noscript," but it does have alternatives that work well.

Re:So who won?

[gl4ss]

it's an advert for views from toms hw, so the article is laid over 17 pages. chrome "won" on windows and safari on osx.

i'm going to say again that the reason why google created chrome was to get a browser that doesn't have adblock for googleads by default. the 40 tabs test is stuuuupid, I reckon it just tests if the browser does some lazy loading or not.

Re:So who won?

[Dyinobal]

Weird I seem to have Adblock and NoScript on Chrome, and on Firefox as well.

Re:So who won?

[beelsebob]

What, you mean Safari? I'm running both adblock, and a JS blocker there.... 'nuff said.

Re:So who won?

[beelsebob]

http://chromeadblock.com/
http://safariadblock.com/

Yep, definitely no adblock there.

Re:So who won?

[Joce640k]

Does the AdBlock really work yet or does it just hide the images once they've downloaded?

Re:So who won?

[Anonymous Coward]

Actually only one of them has Live Bookmarks. That's the one I'm using. I'd switch if the others did.

Re:So who won?

[Anonymous Coward]

Yeah, I'm getting tired of the Tom's articles recently that seem to blatantly want me to click through to lots of pages. Tricky headlines with a mediocre article (although this one was interesting).

Re:So who won?

[Dyinobal]

No idea, but I don't see the ads so I it seems to function just fine regardless of what method it uses.

Re:So who won?

[thetoadwarrior]

Quit visiting shitty sites and you don't need those things.

Re:So who won?

[Joce640k]

You mean like Slashdot?

The page you're reading has scripts loaded from doubleclick.net, addthis.com and googleanalytics.com.

Re:So who won?

[_0xd0ad]

it's an advert for views from toms hw, so the article is laid over 17 pages

No shit. And then it pissed me off even more when I discovered that you have to register to get the printer version.

Re:So who won?

[thetoadwarrior]

You're also using it for free so the least you could is give them ad revenue and disable 3rd party cookies for tracking purposes and as an added benefit they recognise you're being a decent visitor and allow you to temporarily disable ads. Not that I've even used that because I'm not so against paying for things that the mere thought of an ad on the page ruins the whole experience for me.

Re:So who won?

[nschubach]

I haven't read up on it lately, but I thought the chrome adblock would not prevent the ad from being requested/downloaded. It would only hide it.

Re:So who won?

[nschubach]

Part of the point of blocking ads is loading performance and privacy.

Re:So who won?

[Sporkinum]

Not only that, Firefox has decent cookie management as well. Nothing else compares for the whole package.

Obviously IE and Chrome don't want to do that, as it cuts into ad revenue. It doesn't make much difference though, as most users go with browser defaults anyway.

Re:So who won?

[wsxyz]

If you just jump to page 3 of the 17-page extravaganza, they talk about Safari's "Reader" feature, that strips out all of the ads and combines all of the click-through pages into one document. It works great on TH's own article.

Re:So who won?

[_0xd0ad]

Yeah, I liked that little bit of irony too. Although I'm sure it still didn't strip out any of the fully unnecessary screenshots. I don't need to see his desktop, much less on Windows and OS X.

and combines all of the click-through pages into one document

Does it really? I was under the impression it just stripped out the ads.

Re:So who won?

[macshome]

Yes, it combines everything into one long document as well if it can. I used to to scroll through TFA.

Re:So who won?

[cain]

From the Adblock Chrome website:

New in version 2.0: Ads are blocked from downloading, instead of just being removed after the fact!

Re:So who won?

[__aaqvdr516]

Yeah, Opera only runs Adblock and Notscript (not a typo). That's too bad, as it's totally not noscript....hey wait a minute..

Nuff said?

Re:So who won?

[DaVince21]

Only one of them runs adblock and noscript...

Maybe two years ago. ;)

You're certain it's an anonymous reader ?

[Anonymous Coward]

Seriously, not a single clue in TFS tells you that the submitter is the author of TFA ...

Re:You're certain it's an anonymous reader ?

[johnsnails]

If thats the case then the author was deceptive how they posted the summary, then again, who would read an article in full and post the results other than the author of said article, so i guess in was kind of a give away :P

My personal experience with Safari/OS X

[Anonymous Coward]

Safari 5.1 on OS X on my 2.26ghz C2D laptop starts up, loads, renders and navigates pages notably faster than any other browser does on my unburdened Win7/64-system that runs a 3ghz C2D. While there is a thing or two that makes me prefer Chrome, Safari under OS X is definitely the absolutely fastest and swiftest browsing experience around.

Re:My personal experience with Safari/OS X

[johnsnails]

yep and vi is unanimously better than emacs :P Disclaimer: I'm not trying to troll!

Re:My personal experience with Safari/OS X

[SquirrelDeth]

Safari is definitely the fastest till your battery bulges from the stress of 40 tabs open in Safari. And IE 1.0 was definitely the fastest browser ever on my 225 MHZ Pentium Pro with 256 MB ram with Voodoo 2 graphics and SCSI hdds all running on Win 95c. (And Slackware)

Re:My personal experience with Safari/OS X

[Anonymous Coward]

exactly how would your battery bulge from a high memory use? i have used safari on osx for over 6 years now, and i know for a fact, since every version of safari, every version of osx since 10.3 to 10.7, that safari doesn't superfluously process content of unfocused tabs. in fact, it sits idly at a mere percentage or two of cpu consumption when you have just a non-intensive tab open alongside dozens of other tabs.

Re:My personal experience with Safari/OS X

[Yamioni]

Once your RAM runs out of space it has store those extra electrons somewhere, so the first place it looks is the battery. Store enough electrons there and the thing starts to bulge. Seriously, this CS 101 stuff.

I'll save some time for everyone

[Lieutenant_Dan]

Windows 7:
1. Chrome
2. Firefox
3. IE9
4. Opera
5. Safari

MacOS (Lion):
1. Safari
2. Chrome
3. Opera
4. Firefox

Safari on MacOS is almost as fast as Firefox on Win7.

Re:I'll save some time for everyone

[kangsterizer]

Its "interesting" to see that browsers are struggling to be as fast as windows on osx.
of course, linux on all tests is missing which is a shame

Re:I'll save some time for everyone

[gilesjuk]

They ran their tests on a hackintosh from what I can see. They mention the OSX tests were not run on Apple hardware.

Now that probably doesn't make a huge difference, it just depends how good the hardware is and how they've make the hackintosh version of OSX.

Re:I'll save some time for everyone

[Anonymous Coward]

Windows 7:
1. Chrome
2. Firefox
3. IE9
4. Opera
5. Safari

MacOS (Lion):
1. Safari
2. Chrome
3. Opera
4. Firefox

Safari on MacOS is almost as fast as Firefox on Win7.

how about some Linux baselines???

Re:I'll save some time for everyone

[Yamioni]

They would have tested them on Linux too but Tom couldn't figure out how to install it.

Re:I'll save some time for everyone

[AmiMoJo]

Chrome has odd performance issues on Windows 7. It seems to be related to Javascript and causes slow scrolling in heavy apps like Gmail, Google image search and Google Reader. These kinds of benchmarks don't show deal breaking problems like that. If reading RSS feeds wasn't so painful in Chrome I'd switch.

You would expect Google Chrome to work very well with Google apps, but I guess not.

Speed? What about plugins?

[aaaaaaargh!]

Honestly, do so many people have such a slow computer that they have to care for such minor speed differences?

I use Firefox because it has so many extensions and plugins. With just a few additional Firefox extensions installed I'm able to run TOR at the click of a button, block Flash selectively, block referer URLs, block Javascript selectively, block "Like" buttons and crap like that, delete cookies and Flash cookies, block Google analytics, control SSL certificates and being warned of bogus ones, and so on. Unfortunately, such functions and tools are essential nowadays. Not to speak with all the non-privacy related plugins available.

Re:Speed? What about plugins?

[royallthefourth]

Honestly, do so many people have such a slow computer that they have to care for such minor speed differences?

Seriously. I just switched from a 95 watt Phenom II x4 to a 45 watt Athlon II x2 because I wanted a quieter computer that wasn't warm to the touch.
As it turns out, it's not only less wasteful but I can't even discern any performance decrease.

Re:Speed? What about plugins?

[fermion]

I suspect these speed differences are in support of ad supported web sites. Something like /. renderes quickly as long as the google ad servers are responding. Something like the NYT renders quickly in my browser, but I suspect that is because Flash is turned off.

If browsers are fast, and don't consume many resources, the user will let the ads run. If the browsers are slow, then users will begin to figure out how to fix the problem, perhaps by blocking ads. It is any wonder why Chrome and Safari are fast browsers? Google and Apple both depend on users allowing the full web experience, at least on full powered computers.

Re:Speed? What about plugins?

[_0xd0ad]

Yeah, I rather enjoyed the graph for page load time labelled "lower scores are better" with Tom's Hardware coming in dead last. I don't think that's what they meant to show by that particular test, but it was amusing.

Re:Speed? What about plugins?

[shutdown -p now]

Honestly, do so many people have such a slow computer that they have to care for such minor speed differences?

Netbooks? Nettops?

I use Firefox because it has so many extensions and plugins. With just a few additional Firefox extensions installed I'm able to run TOR at the click of a button, block Flash selectively, block referer URLs, block Javascript selectively, block "Like" buttons and crap like that, delete cookies and Flash cookies, block Google analytics, control SSL certificates and being warned of bogus ones, and so on. Unfortunately, such functions and tools are essential nowadays. Not to speak with all the non-privacy related plugins available.

Similar plugins are available for other browsers. I'm not sure if the list is as extensive, but blocking JS and ads is certainly there.

And, frankly, 99% of users only care about blocking ads. The rest on your list is geek-centric.

Re:Speed? What about plugins?

[Anonymous Coward]

Us geeks never feel it, because we never have underpowered systems; unlock the admin accounts and hand the same PC to your John Smith friends, and go back in a year. If you didn't install an antivirus and provided it to a pr0n watcher who uses IE, there's bound to be a problem or two, even if you have 6GB RAM.

But most mainstream users of fullsize home desktops out there still have pirated XP systems and 500MB - 2GB RAM unless they really went out to buy the crappy Vista machines in 2007. Oh, and they're running single cores.
That said, when I bought this quad core to upstage my slow solo desktop and duo laptop, I didn't notice speed improvements in daily explorer use and launch times so much. What I did notice is that actions just stopped stuttering.

Stutter is the sign that you need more RAM and other things, but John Smith just thinks their PC has literally fallen sick and he needs a new PC.

Re:Speed? What about plugins?

[Anonymous Coward]

Must be running windows

Still looks like they're all broken to me

[Anonymous Coward]

It still looks like every web browser is broken to me.

For example the first place Chrome on Win 7, has shitty history management.

quote "...As a PC guy..."

[Anonymous Coward]

Do people still talk like this?

The best thing about Reader in Safari...

[pknoll]

is that it works on Tom's Hardware articles.

Not mentioning all factors

[Pooya_M]

Firefox awesome Add-ons and being open-source are not mentioned in this comparison...

Re:Not mentioning all factors

[m50d]

If they'd mentioned the add-ons then they should've run the test with a few installed; my guess is it'd make firefox's performance even worse.

Re:Not mentioning all factors

[Pooya_M]

EFFICIENT add-ons are more valuable than tons of DUMP so-called add-on.BTW, Firefox also runs on Linux based OS

Re:Not mentioning all factors

[Anonymous Coward]

Being open source is hardley relevant to a browser feature comparison. At least in the context of this article.

But then, I don't subscribe to the RMS dogma.

Re:Not mentioning all factors

[m50d]

EFFICIENT add-ons are more valuable than tons of DUMP so-called add-on.

Yeah, and a browser that has the right features in the first place is more valuable still.

BTW, Firefox also runs on Linux based OS

And? So do most of the browsers they mention.

Re:Not mentioning all factors

[Anonymous Coward]

I'm from Croydon, but thanks for the generalizations.

I don't think open source is useless, bad or anything else. I just this it's a useless comparison in the context of this article. That said, I prefer the BSD license.

FF6 + Hardware Acceleration = BSOD

[Stavr0]

Just like in a Grand Prix, pushing beyond the limits results in a blown engine. Since update to ver 6, I've had daily BSODs that are caused by a video driver crash. Since I turned off hw acceleration in FF6, no more BSODs.

Re:FF6 + Hardware Acceleration = BSOD

[_0xd0ad]

There is no reason a user-level application should be able to cause a system driver to crash. Sounds like you have a shitty driver. Why are you blaming it on Firefox?

Re:FF6 + Hardware Acceleration = BSOD

[Anonymous Coward]

Because only firefox makes it crash?

Re:FF6 + Hardware Acceleration = BSOD

[shutdown -p now]

It doesn't matter. Any crash of this kind is still a bug in the driver, no matter which application has the bad luck of triggering it with its usage patterns.

No Linux? Bah.

[serbianheretic]

No testing under Linux ... like it is 1999. And this is on supposedly geek site? Meh.

Re:No Linux? Bah.

[Anonymous Coward]

If you actually read the article, you'd notice that their last test was Windows versus Linux, and this one was specifically in response to people asking that Mac OS X be compared.

Re:No Linux? Bah.

[MonsterTrimble]

Exactly. Plus with Opera getting beat by both Firefox and IE9 on Win7 I call bullshit.

Re:No Linux? Bah.

[Kjella]

And much like 1999, the Linux web browsing desktop has less than 1% market share. There's absolutely no imaginary numbers of hidden users you could pull up that would make them more significant in a browser test. There was a time when I wanted to see a trend, but it's not there so you can't really call them up and coming either. Of course it hurts here on slashdot but the truth is pretty plain to see.

Re:No Linux? Bah.

[Anonymous Coward]

> blablabla

Linux or FuckYou

08/15-users are not in the audience.

Re:No Linux? Bah.

[Intropy]

The latest numbers I saw at Gartner had Linux at about a 2% market share. Compare that with OSX's 4%.

Re:No Linux? Bah.

[Anonymous Coward]

No testing under Linux ... like it is 1999. And this is on supposedly geek site? Meh.

They mentioned in TFA that they'd done a Linux comparison in a previous "grand prix" article. And the Linux browsers generally lagged Windows on the same hardware.

That was some versions ago, though.

Moot though, since practically no one uses Linux as a desktop OS for browsing. There are probably more Android browser users than Linux desktop/laptop users.

Re:No Linux? Bah.

[ryaxnb]

Opera has been slow recently because it's javascript engine lags behind firefox and particularly chrome. Opera has a relatively fast plain html renderer that slows when encountering Javascript moreso than other browsers including IE 9 which has a heavily tuned javascript engine (unlike IE 8). As for link desktop share it is generally agreed to be about 2% or 1/4 the share of Macintosh.

Re:No Linux? Bah.

[bigpet]

Here's the thing with making benchmarks under Linux:
People will have something against your methodology no matter what. "X is faster than Y under windowing toolkit Z" or "No wonder X beat Y when you didn't even install kernel patch Z" or "You better retest this on kernel version X because Y is known to ..."

Sure you can give ballpark estimations on a specific setup but that's about it.

Re:No Linux? Bah.

[dudpixel]

Who cares about market share? Its a geek site, so linux is not outside their scope by any stretch of the imagination. OSX only has 9-10% doesn't it? and yet they include that without any quibble about market share.

It appears they did include linux in earlier benchmarks (this article is Web Browser Grand Prix 6, and there were 5 others before it).

I'd say the complaint about linux missing is valid...however I'd say they would have used ubuntu due to its popularity and maybe the linux crowd would still complain that it wasn't fedora, or opensuse, or gentoo...

Re:No Linux? Bah.

[jaminJay]

Market share is slowly trending up in an ever-expanding market. 1% is still many millions of users. It's not as though the results would be that much different, causing people to flood to the platform, surely?

Re:No Linux? Bah.

[Kjella]

The latest numbers I saw at Gartner had Linux at about a 2% market share. Compare that with OSX's 4%.

Source please.
Hitslink: 0,91% vs 5,61%
Statcounter: 0.77% vs 6.27%

If you saw anything like 2%, it probably includes Android which has somewhat over 1% of the browser market, most of which use the built in browser. At least "What's the best browser for an Android phone?" is a completely different test...

Re:No Linux? Bah.

[Intropy]

Good catch, the page I saw referenced this: http://www.gartner.com/it/page.jsp?id=1762614. So reading that, the figures are estimates for end of 2011 not current values. It also has OSX at 4.5% and Linux at "below 2 percent." Desktops, laptops, and notebooks were included. Phones, tablets, and the like were excluded.

Re:No Linux? Bah.

[MonsterTrimble]

Interesting. I never use Firefox (due to it historically having extremely poor performance on my machines) and IE9 (work computer has IE8 and I use Linux at home). I haven't noticed any slow down on Opera on my linux boxen. I do have chromium installed and it is faster but I like the built-in RSS feeder Opera has as well as the file sharer.

Re:No Linux? Bah.

[neonsignal]

I can't believe that a site like Tom's Hardware is scared of a little criticism. And it seems obvious that you would do the benchmark on the current Ubuntu release, if only because of its relative popularity.

Any of these systems can be tweaked, not just Linux; if that was an argument, then no-one would ever benchmark anything.

Firefox 6???

[Urza9814]

Firefox 6? C'mon! I'm already on Firefox 7! Oh wait, hang on, there's an update for Firefox 8 now. Or should I go with 9 beta? Eh, 10 should be released tomorrow, right?

Re:Firefox 6???

[cgraeff]

This joke is getting tiresome.

Re:Firefox 6???

[Anonymous Coward]

Exactly, you could say the same about Chrome, there was 11 just 2 months ago.

Re:Firefox 6???

[Anonymous Coward]

this ceased being funny several weeks ago. who gives a fuck about firefox version numbers

Re:Firefox 6???

[schnikies79]

Not as tiresome as their release schedule.

Re:Firefox 6???

[Anonymous Coward]

So are the versions.

Re:Firefox 6???

[Anonymous Coward]

where's the chrome release version love? I'm currently on Chrome 15, and it came out plenty of years after firefox....

Re:Firefox 6???

[gilboad]

[RANT="Sorry for being blunt, but the continuous stream of stupid Firefox/Linux kernel version jokes/rants is *really* getting on my nerves".]
As far as I remember, chrome uses a somewhat shorter six week [1] release schedule. (Feel free to correct me if I'm wrong).
Now given the fact the in one of you previous post, you said: "... I'd still be fine with it, but I'd probably be migrating entirely to Chrome soon" [2], I'm left to wonder if you plan to make the same tedious jokes about chrome...[/RANT]

[1] http://en.wikipedia.org/wiki/Google_Chrome#Release_channels_and_updates
[2] http://slashdot.org/comments.pl?sid=2273874&cid=36584518 (Actually took the time to look it up)

Re:Firefox 6???

[p0p0]

That's true, but at the same time NONE of my Chrome extensions broke or disabled unlike what happens after every Firefox update.

EVERY. SINGLE. UPDATE.

Re:Firefox 6???

[Urza9814]

Wow. You're really getting annoyed by all the Firefox release cycle jokes, huh? :)

So, first, I said I WOULD be migrating to Chrome IF Firefox decided that corporate users were of more importance than home users. But I run Linux -- where else would I go? I suppose I could switch to Opera, but I dislike them for numerous more significant reasons. In my mind, Firefox and Chrome are by far the best browsers for Linux -- which one I use at this point is just a matter of which one has less minor annoyances. Right now it's definitely tipping towards Chrome, though not enough to get me to alter my habits and swap all my saved tabs. Besides, I use Chrome for development, where I really need a good browser; Firefox is just for checking my mail, news, etc.

The thing about Chrome is that it updates well. I've never had my download history corrupted causing a crash of Chrome every time I download a file (happens with EVERY Firefox update.) I don't get prompted if I want to kill the browser immediately to update with Chrome, it does it silently. I don't get a 'YOU JUST UPDATED CHECK OUT ALL THE COOL NEW FEATURES!!!!' tab popping open every time Chrome updates. I don't get a 'You just updated, gotta check your plugins -- oh no, none of them are compatible!' window with every Chrome update. Basically, it doesn't matter how often Chrome updates, because you don't notice that it even does it. With Firefox, it's extremely noticeable. Not that I really mind ALL of these things, it would be somewhat nice to get a single 'oh hey, BTW, I just updated myself' notice, but Firefox started doing short release cycles before they managed to make it easy and seamless to upgrade.

You can't joke about a short release cycle if you never notice there's even been a new release.

Re:Firefox 6???

[gilboad]

The thing about Chrome is that it updates well. I've never had my download history corrupted causing a crash of Chrome every time I download a file (happens with EVERY Firefox update.) I don't get prompted if I want to kill the browser immediately to update with Chrome, it does it silently. I don't get a 'YOU JUST UPDATED CHECK OUT ALL THE COOL NEW FEATURES!!!!' tab popping open every time Chrome updates. I don't get a 'You just updated, gotta check your plugins -- oh no, none of them are compatible!' window with every Chrome update. Basically, it doesn't matter how often Chrome updates, because you don't notice that it even does it. With Firefox, it's extremely noticeable. Not that I really mind ALL of these things, it would be somewhat nice to get a single 'oh hey, BTW, I just updated myself' notice, but Firefox started doing short release cycles before they managed to make it easy and seamless to upgrade.

You can't joke about a short release cycle if you never notice there's even been a new release.

I must admit that -my- experience (anecdotal, just as yours) is the exact opposite of what you are seeing.
I -never- had history/bookmark/etc corruption due to version upgrade on Firefox, while my experience with Chrome is the exact opposite (Addons no longer working; crashes; etc). I would seriously consider switching distribution to one that's **apparently** more competent in packaging Firefox...
As for the part about having the "new version" tab, I can't really get the point: would it make it any easier on your eyes if the version was 3.8 instead of 6? (Though I would agree that Firefox devs should do their best to help addons developers write version-free addons)

Much like the Linux kernel version issue, users seem to make far, far, far, far, far (x10000) too much fuss about mere version numbers.
Version numbers are meaningless, get over it! (Did Microsoft really release 1996 NT versions/builds between NT 4.0 and Windows 2000?)

- Gilboa

good mourning, tired of terror monday here again

[Anonymous Coward]

evidenced by the dogs volunteered to go outside today. had to shave with pliers again (like monkey hair)? some woman down the road a bit got shot (6 times) while brandishing a pellet gun. tell your kids what can really happen now. plans are moving ahead to have the fallen gargoyles (dc & ny) re-installed? so no stuff that really matters to be concerned about yet, again, today.

thanks to those who have been disarming, & reading the teepeeleaks etchings. hairy stuff to say the least.

Maze solver. bah!

[Anonymous Coward]

As a long time firefox user, I am quite surprised, how it underperforms in the CSS performance test, maze solver - http://ie.microsoft.com/testdrive/performance/mazesolver/default.html
While chrome can do that in a jiffy!

Re:Maze solver. bah!

[royallthefourth]

I imagine a test designed by Microsoft to make IE look good might have also been designed to hit Firefox's weak points. Maybe it's slow at that demo, but it works great for my daily browsing. What else should I care about as far as speed goes?

Re:Maze solver. bah!

[BZ]

This test exercises a situation that's very rare on the web (where by "rare" I mean that it's only been encountered in this test to my knowledge): thousands of absolutely positioned elements that are all being moved around using CSS transforms, with each one only being moved once by going from no transform to a translate transform. That's just not something anyone other than this test does. Most people who want to move an absolutely positioned element just change its .top and .left, but this test sort of went out of its way to do things the weird way.

The net result is that this test ends up hitting a rare-case O(N^2) codepath in Gecko. See https://bugzilla.mozilla.org/show_bug.cgi?id=641340 and https://bugzilla.mozilla.org/show_bug.cgi?id=641341 and https://bugzilla.mozilla.org/show_bug.cgi?id=670311 for the bugs tracking this on Mozilla's end.

Fixing these has not been a terribly high priority, since it would mostly affect this one synthetic benchmark (I say "mostly", because bug 670311 could have benefits elsewhere too).

I wanted to see beta's

[Artem S. Tashkinov]

It's a bit sad they haven't tested betas of Firefox, IE and Chrome.

E.g. Firefox 7 includes some memory usage optimizations which could easily halve its memory usage under the stress test Tom's Hardware guys carried out.

Re:I wanted to see beta's

[PortHaven]

Um, they're betas...

Guess they should be testing IE10 beta as well.

Re:I wanted to see beta's

[shutdown -p now]

Chrome 13 and Firefox 6 are both released, they're not betas.

Browser benchmarks under Linux

[microphage]

> This is one of the better 'browser battle' articles, though. Chrome 13, Firefox 6, IE9, Opera 11.50, and Safari 5.1 are put through 40-some tests on both Windows 7 and Mac OS X Lion ..

Where are the browser benchmarks under Linux?

Re:Browser benchmarks under Linux

[gullevek]

Did you read the article? This was a shoot out between OS X and Windows. If you would have read it, that there is another one done before that compares Windows 7 and Linux.

WTF is with the false statement about Safari/Mac

[_0xd0ad]

Every one of these features is completely unique to Safari for Mac. ... Like Lion itself, Safari 5.1 supports several multitouch gestures ... Swiping upward with two fingers causes the page to scroll down. Likewise, swiping downward with two fingers scrolls the page up. ... Using the same two-finger swipe as the scroll gesture, performed left and right, controls navigation. Swiping two fingers to the right navigates to the previous page in your history, and swiping left moves forward.

That's complete bull. My HP laptop supports multitouch gestures just fine, with the exception of the Mac's gestures all being exactly backward. Swiping with two fingers scrolls (in any of the 4 directions, not just vertically). Swiping with three fingers navigates forward/back.

Chrome cant even scroll Google Image Search smooth

[Jackie_Chan_Fan]

Chrome just has terrible page rendering performance. Firefox scrolls so much smoother.

The best comparison is google image search. Chrome can not even scroll google's own image search smoothly. Firefox does it as smooth as butter. Chrome also scrolls bing's image search poorly as well. Firefox wins in rendering performance.

Even the fish bowl test shows firefox is far better at rendering.

Benchmarks benchmarks benchmarks

[kangsterizer]

As usual, benchmarks are quite broken.
I can understand when you get a few FPS on a graphic card but.. lets take startup time..
0.8s vs 1.1s and the bar is like so much bigger, while in reality this makes almost no difference.

Then again, testing stuff such as acid3 (which will never be implemented in some browsers to reach 100% because of things the acid3 did not foresee) or memory release wrongly (per process model is forced to release all the memory, threaded models keep a lot in the cache)

there's many other such cases, which basically mean, you can put any of the contenders in the top place (except safari on windows i guess!)

Palemoon is a faster Firefox

[Anonymous Coward]

http://www.palemoon.org

Just tossing it in there. Firefox isnt as fast as it should be.

Tech marches on....

[Anonymous Coward]

Okay, we're no longer in the days where your 486SX with 2MB of RAM would suffer greatly if your code was interpreted versus compiled in hand-crafted ASM.

For the average user who has more RAM and CPU cycles than what they know what do with; does it matter...

Yes it matters, because we're no longer in the days where websites are static HTML and images. You might wish it were so, but tough cookies.

Oh, and 640k of memory isn't enough for anybody these days.

Re:Tech marches on....

[Skarecrow77]

Yes it matters, because we're no longer in the days where websites are static HTML and images. You might wish it were so, but tough cookies.

Sure it is, if you have the right browser extensions. Hell, I browse from work with images turned off completely. The web is so responsive.

Oh, and 640k of memory isn't enough for anybody these days.

Ever looked at the demo scene from the early 90s, and seen what those guys could do with just 64k, let alone 640k?

The reason that 640k isn't enough for anybody is that once they found out they had more than 640k, they said "well hell, might as well use it". We've gotten to absurd points where tray icons and such are eating up 30, 40, or 50mb of memory, just because "hey, it's probably there". These coders are using a dump truck to move a deck of cards just because they have access to a dump truck.

although having access to a dump truck IS pretty cool.

And the loser is. Tom's Hardware

[Stonent1]

Notice in the benchmarks the absurdly long load times for Tom's Hardware compared to other sites (however not much different from Huff-Po) There's so many ads and crap coming from 3rd party servers that depending on which ad you get, it can severly impact the load time.

Re:Uninformed fanboi much?

[Anonymous Coward]

FYI: IE9 is actually the only major browser on the market that has a selective JavaScript blocker built-in, as part of its somewhat misleadingly named Tracking Protection feature. With the help of the latter you can block scripts (or CSS files, embeds, whatever) either manually, by specifying an URL, or you can download and use many of the "blocking lists", which actually work much the same way AdBlock Plus' do. And you don't even need to download a plugin (like with Firefox or Chrome) for that to work.

None of my needs were met by their tests :(

[sgt scrub]

I use a process of elimination starting with the most important.

I want to know if the browser is usable on any machine I use. Firefox, Chrome, Opera.
I want to know if the browser is capable of displaying content I access without extensions. Firefox, Opera.
I want to know if the browser is capable of protecting me from certain content. Firefox, Opera
I want to know if I can enhance the browsers abilities. Firefox

Re:None of my needs were met by their tests :(

[Ksevio]

You can create extensions for Opera these days (as well as the user javascript and CSS files), so unless you're modifying the source of the layout engine and building your own, then It looks like your final category still has Firefox and Opera.

Re:None of my needs were met by their tests :(

[sgt scrub]

I was really meaning something deeper in but I didn't know they had extensions. If you use XMLRPC to enhance SVG Opera thinks your displaying an animated image and spins continuously. It also doesn't handle cursor assignments when you "use" "symbols". Neither imposition is worth them fixing any time soon.

For example, the code below moves an image of an arrow to a specific position in an SVG based interface. When displayed by Opera, the cursor doesn't change to the pointer.
Note, I had to remove the greater signs for the code to display on /. It would be nice if there was a [code] tag.

symbol id='uarrow' viewBox='0 0 32 32'>
path style='cursor: pointer;' fill='green' stroke='green' stroke-width='0.1' d='m16,0l-16,32l32,0l-16,-32z' onclick='moveDetails(0)'/> /symbol>
use x='780' y='450' width='32' height='32' fill-opacity='1' xlink:href='#uarrow'/>

Its nice to know they allow extensions. I'll have to check and see what they have for VRML and XUL. Thanks!

Re:None of my needs were met by their tests :(

[swilly]

What content can you not display in Chrome without an extension that you can in Firefox?

What content will Firefox and Opera protect you from that Chrome won't? As far as I know their content protection is about equal, and Firefox and Chrome us the list of malware to block provided by Google.

Both Opera and Chrome have extensions. Maybe not the ones you want, but all three browsers allow you to enhance their abilities. And it is worth pointing out that Chrome is rapidly catching up to Firefox in this area.

I made the switch from Firefox to Chrome and I couldn't be happier. The only thing I miss is the FlashGot extension.

Re:None of my needs were met by their tests :(

[_0xd0ad]

Note, I had to remove the greater signs for the code to display on /. It would be nice if there was a [code] tag.

Slashdot allows you to format your posts with HTML. Use &lt; for <. And you removed the less-than signs, not the greater signs.

Re:None of my needs were met by their tests :(

[dudpixel]

I use a process of elimination starting with the most important.

I want to know if the browser is usable on any machine I use. Firefox, Chrome, Opera.
I want to know if the browser is capable of displaying content I access without extensions. Firefox, Opera.
I want to know if the browser is capable of protecting me from certain content. Firefox, Opera
I want to know if I can enhance the browsers abilities. Firefox

this seems a bit inconsistent don't you think?

Surely criteria #4 makes #2 irrelevant, if #4 satisfies #2 (minus the "without extensions" bogus requirement).

It reads like you actually started with Firefox as your answer and then constructed the criteria afterwards...

Re:None of my needs were met by their tests :(

[sgt scrub]

lysdexia kills

Re:None of my needs were met by their tests :(

[sgt scrub]

It was process of elimination. Chrome doesn't sufficiently do SVG for me. (note the "my needs" part of the Subject). I don't doubt they will catch up, and probably even support SVG 1.2 before the others. If they supported textArea they would have bumped Firefox. Well, if they didn't still have a problem with POST. I will never send user:pass in the URI or cookie.

Re:None of my needs were met by their tests :(

[sgt scrub]

If you look at it you will see it is ordered by work. It starts with "can I install it and be on my way" and ends with "will I have to rewrite base code to get what I need". #3 is such an important need it is listed. It is also the reason "without extensions" is important. For example, adding flash to a browser isn't the safest thing one can do. HTML5 canvas, despite its potential vulnerabilities, can replace that issue.

Re:None of my needs were met by their tests :(

[dudpixel]

That use case (no flash) isn't satisfied by any of those points.

I believe you can disable flash in chrome, if it bothers you. The linux version doesn't include it so its a non-issue there too.

Not that I'm saying you should use chrome, or any other browser. It just reads exactly like you started with Firefox and thought up a set of criteria that only Firefox can satisfy.

How do I upgraded laptop?

[tepples]

Upgrading to 8 GB would require buying bigger RAM modules than this computer's motherboard recognizes. So it would mean buying a new motherboard, which in the case of a laptop would usually mean a whole new computer. Furthermore, 64-bit Windows and low-volume peripherals do not mix due to the annual cost of KMCS certificates.

Flashblock

[tepples]

I'll give a site ad revenue as long as it chooses to show text or still-image advertisements. I've just blocked the MIME type of a certain Adobe product for sites not on a whitelist. Do you think add-ons like Flashblock are an acceptable compromise between the interest of keeping my web browser experience clean and fast and site owners' interest of drawing ad revenue?

Cap

[tepples]

I don't see the ads so I it seems to function just fine regardless of what method it uses.

If you were on a 5 GB per month Internet plan, you'd probably care more.

Poor Sad Sap

[PortHaven]

Just wait, once those advertizers dump Flash for HTML5. You'll have no way to block those ads. And even slower performance. Than you'll wish for those days when complex visuals and animations were done in a plug-in.

"Be careful what you wish for..." ;-)

Re:Poor Sad Sap

[_0xd0ad]

Just wait, once those advertizers dump Flash for HTML5. You'll have no way to block those ads.

Sure you can. Block the <canvas> tag with element hiding rules and/or block the Javascript that makes it run.

Re:Poor Sad Sap

[ctrl-c_ctrl-x]

And , and . If all you want is text use an alternative text-based browser. If you don't want rich unique professional content, block ads and trcking. The web as we know and love it, with all the great applications and sites and media would not exist without ads. We tried the free and the subscription business models, and for the most part it didn't work. View the ads and enable the tracking and the web will continue to advance. If those annoying flash and pop-up ads hurt traffic so much, sites would not use tem because noone would buy them. Complain all you want, but as long as people want to be paid for the content they produce, ads are here to stay in all their flashy, animated glory. Stop being such a pretentious Luddite.

Re:WTF is with the false statement about Safari/Ma

[Andreas Mayer]

with the exception of the Mac's gestures all being exactly backward. Swiping with two fingers scrolls (in any of the 4 directions, not just vertically). Swiping with three fingers navigates forward/back.

In Lion swiping with two fingers does both. It scrolls to the end of the view and then it navigates to the next/previous page.

Actually quite nice.

Nonsense

[Anonymous Coward]

Testing browsers on hackintosh is like testing a ferrari with vespa wheels.

The Sad Denial

[PortHaven]

Let's admit it folks. We're in a bit of denial. Sure,

Chrome came out on top. Still not sure why it's not my default browser. And why I cling to Firefox. I think it's cause more sites are still tested with Firefox.

Safari, showed it's true colors (I've always found Safari on Windows to be slow and fail to load pages. These benchmarks just seemed to confirm my feelings.) I've not been a fan of Safari. Even on Mac it was my secondary browser.

Firefox has been growing slower, and slower. And these tests just seem to confirm what we already know.

Opera, hmm...the little browser that's still there. Why has Opera never taken off. Not really sure.

But let's talk about the dirty little secret. IE9. How many tests was IE9 at the top? A lot...in fact, it and Chrome were doing most of the battling in performance. IE pretty much lost in one area. And granted to many that's an important area. But the truth is, IE is really making some big gains.

My point being, is if further improvements to IE10, particularly in standards compliance, are made. Will we be honest without ourselves and give it it's due?

I've had to start admitting that IE9 is loading many pages faster than Firefox for me. It's a hard pill to swallow.

Re:The Sad Denial

[ctrl-c_ctrl-x]

IE > 8 is making great progress. I just wish that it worked on non windows boxes.

Re:The Sad Denial

[jo_ham]

Safari is pretty good on the Mac - I used to use it alongside Firefox but switched to Safari/Chrome after the whole status bar debacle and the trainwreck that was FF4 (and it looks like I got out at the right time).

I know that I've essentially got two Webkit browsers now, but I'm much happier. In terms of speed I don't really notice any difference in real world use between Safari and Chrome except that Safari has a big memory leak when using Adblock, so it eventually swallows up all the available RAM you have and you need to restart it. It doesn't do this without the extension, so it's something to do with that but it's too useful to leave it uninstalled.

Re:WTF is with the false statement about Safari/Ma

[_0xd0ad]

Ugh, no thanks. It's bad enough that scrolling in text boxes does that sort of thing: scrolls to the bottom of the text area and then starts scrolling the page.

Bullshit

[pavon]

I have two computers. A Windows 7 laptop with 4GB of memory and a Debian desktop with 3GB (maxed out). Both computers get unusably slow after a few days. Only then do I check the resource usage and find that Firefox has managed to grow to consume 70-80% of the physical memory. Worse, since it since it constantly touches all it's memory pages, it is the other applications that get pushed out to swap, not Firefox. I restart Firefox, and it's memory usage drops to a 1/3 of what it was using, with the same windows open.

Now I don't know exactly what is causing the memory growth. It could very well be memory leaks in the JavaScript that webpages are serving, and not in Firefox's code. However, they are the ones pushing for rich web applications, so they need to take responsibility for managing these rich applications, just like an OS does now.

Re:Uninformed fanboi much?

[b0bby]

Yeah, TPLs are good - the first thing I do now on any machine I'm messing with for family is add fanboy's ie9 tpl.
http://www.fanboy.co.nz/ie.html

Re:WTF is with the false statement about Safari/Ma

[jo_ham]

You can, of course, turn that off. You can also customise all the gestures to your liking, so if you feel it is "exactly backward" you can swap it over.

Re:WTF is with the false statement about Safari/Ma

[_0xd0ad]

Ok. All I'm saying is that multitouch gestures are not at all unique to Mac or Safari.

Re:Chrome cant even scroll Google Image Search smo

[Billly Gates]

"Chrome just has terrible page rendering performance. Firefox scrolls so much smoother."

Want an even bigger surprise? Use IE 9 and then scroll and see how smooth it is? IE 9 has Chrome and Firefox beat with GPU accelerating browsing and these benchmarks only focused on load time and not the actual experience of rendering.

On my 3 year old laptop IE 9 is not that much better, but my desktop has an ATI 5750 GPU and I noticed the difference. /.ers typically ignore anything IE these days and pretend it doesnt exist.

Browser preference.

[m1ndcrash]

Even though a lot of sites / online services are built primarily for IE, I don't use it as it undermines overall security. In fact I use the combo of Opera-Chrome-Firefox. Opera is my day-to-day working horse browser. It has less 0-days on it's record and seems to be more convenient for casual browsing. Chrome is for Silverlight use and Firefox is for rendering issues.

Re:The Sad Denial - IE 9

[Billly Gates]

This is slashdot. They will acknowledge it even exists except at work as a silly PHB browser that they need to support.

I have been flamed and modded down before when mentioning poor rendering performance of Chrome vs IE 9 on a decent GPU. The fact of the matter is IE 9 won Flash and HTML 5 rendering speeds. Sure Chrome loads faster but, try to go to www.msnb.com or Google Images and hit the up and down arrow keys with all 3 browsers?

If you own a Nvida 8600/ATI 2500HD or higher you will see IE 9 smooth with only a little flickering. Firefox is ok but very slow. Chrome is choppy and flickers like mad. On slashdot I can not even read the comments scrolling up and down with Chrome.

On my old laptop they perform about the same but that is because it has an ATI x1200 2007 era.

IE 9 also won ram usage with lots of tabs opened too. Chrome is better but still can take gigs of ram if you open 40 tabs and leave it on for a day or two. IE 9 has selective Javascript blockers too and xss eliminating noScript for most uses ... something most slashdotters whine about and ignore this.

IE brings back bad memories and I do not blame people for shuddering it. It was a decent browser 10 years ago when IE 6 was new as I remember many slashdotters saying they couldn't stand Mozilla/Netscape and that IE was good. It went to hell for 9 years as MS secretly hated it and wanted to return to kill the web and go back to client/server Win32 aps. But MS is coming back out and embracing it now as they realize they have no choice. It will take years to get rid of IE's tarnished image even if it is a decent browser today and is certainly modern again. It is not IE 6 or IE 7 anymore.

NO LINUX!

[Pink_Frosty]

I think the whole thing needs to be thrown out with no Linux tests.

Re:NO LINUX!

[madhi19]

The funny part is the web server used for this benchmark is Ubuntu 10.04 Server Edition. So technically all the test rely on Linux!

Response from epSos.de

[Anonymous Coward]

You will have to use at least two of them, if you are under Linux.
Opera and Firefox for browsing and Chrome for all the flash stuff.

Re:WTF is with the false statement about Safari/Ma

[Anonymous Coward]

Netbooks too. It's more to do with the touchpad and driver than the OS.

yes, yes

[Anonymous Coward]

yes, we know. chrome is the best, firefox sucks. mozilla can't do anything right. blah blah blah.

terrible memory management test

[locopuyo]

Opera's memory management is much much better than their tests show. Opera allows you to set the amount of memory you want it to use and by default is set to automatic which simply takes up as much as you have free. The benchmark Tom's Hardware used simply looks at how much memory is being used and claims more memory used is bad which is a very bad benchmark. Opera uses available memory and reloads pages much faster than other browsers which should really put it in first but according to their test puts it in last.

Re:Chrome cant even scroll Google Image Search smo

[Jackie_Chan_Fan]

IE9 is very impressive in that area. In some areas it is lacking. I found firefox ran some html 5 stuff better.

IE really needs to improve its ui. I'm not a fan of its ui as it is now. It needs better extensions too