Web Browser Grand Prix VIII: Chrome 16, Firefox 9, And Mac OS X

[adamovera]

[citation][nom]pmcoz[/nom]I'd love to see the inclusion of tests to measure the privacy (and bandwidth efficiency) of each browser in terms of phoning home, or anywhere else not directly applicable to the page being reviewed. This would need to include installation, startup and page load tests. This is my major concern about Chrome.[/citation]
I would too. Any idea how to go about doing that? If you like Chrome, but not Google, you can always use Chromium, it's open source and dead simple to install and update on most modern desktop Linux distros - not sure about Win/Mac though.

 

[adamovera]

[citation][nom]mayankleoboy1[/nom]can we have the linux circuit next time?[/citation]
I wanted it this time! Well, back in November when Firefox 8 arrived that is. But the default Linux distro is Ubuntu and version 11.10 had issues getting the most recent versions of all three browsers out in a timely manner. Will keep trying though, I'd say it's probably my top priority for the series right now.

 

[adamovera]

[citation][nom]mi1ez[/nom]it should be ADO. How can professional journalists make such simple mistakes time and time again?![/citation]
Damn, I wonder why the spell check proposed a French word, but it's not as though my original spelling was anywhere near ADO Thanks, good catch! It's fixed now and should update soon.

 

[adamovera]

[citation][nom]rhea[/nom]Any plans to run benchmarks on Linux? I'd be very interested since I use Linux as my main OS and I am sure lot of other Tom's readers do too.[/citation]
Absolutely, see previous two replies on the subject. BTW, what distro are you using?

 

[adamovera]

[citation][nom]pc elitist[/nom]The winner is the user.[/citation]
So true. I might just have to take that line.

 

[joeman99]

well well well , Apple's "elitist" platform is crap compared to windows. i just love how those elitists install windows on their "elitist" devices to have access to real software and not just toys. get that isheeps , safari is slower on macos than on windows. trolololo. speding a lot of money on your hipster devices and you're still loving the pack.

 

[wheredahoodat]

It would be cool if we have have an article like "What browser is best for Grandma or your kids." Then you can look at things like ease of use, security, and then finally personal preference. It would be entertaining to see what browser non tech people would actually use if they had a choice and were exposed the all of them.

 

[madooo12]

whenever anyone funds a test his browser is usually the safest

 

[madooo12]

actually opera supports HardWare acceleration in version 12 alpha, you should test it also

 

[adamovera]

[citation][nom]joeman99[/nom]well well well , Apple's "elitist" platform is crap compared to windows. i just love how those elitists install windows on their "elitist" devices to have access to real software and not just toys. get that isheeps , safari is slower on macos than on windows. trolololo. speding a lot of money on your hipster devices and you're still loving the pack.[/citation]
I think you got something crossed here. Safari for OS X is actually FAR superior to Safari for Windows. The strange part of the whole cross-platform comparison is that the OS X browsers fared better (versus the Windows browsers) on a Hackintosh PC than an actual Mac.

 

[adamovera]

[citation][nom]wheredahoodat[/nom]It would be cool if we have have an article like "What browser is best for Grandma or your kids." Then you can look at things like ease of use, security, and then finally personal preference. It would be entertaining to see what browser non tech people would actually use if they had a choice and were exposed the all of them.[/citation]
Anything that auto-updates will be the safest in terms of security, however, constant unannounced interface design changes that occur in those updates will also cause that crowd some ease-of-use hiccups. You also have the social manipulation (phishing) attacks to worry about more than any 'hacking' or malicious code. These types of scams will almost always get past any code-based security measures. And despite what Microsoft has implemented, it has been my experience that asking people "Are you sure you want to do that?" over and over simply becomes repetitive and normal, and therefore ignored. But I agree that it would be nice to expose that type of crowd to all of them and let them decide in some kind of study. Unfortunately, it's also been my experience that that crowd will most likely always go with what is familiar - probably why IE6 took so damn long to kill off. Great comment, tough issues.

 

[adamovera]

[citation][nom]madooo12[/nom]actually opera supports HardWare acceleration in version 12 alpha, you should test it also[/citation]
The minute Opera 12 goes final, we'll be testing it. I too am looking forward to that release.

 

[JonnyDough]

[citation][nom]hardcore_gamer[/nom]The only one thing I hate about firefox is that it takes a lot of time to launch.[/citation]

May I suggest a solid state disk?

 

[JonnyDough]

[citation][nom]A Bad Day[/nom]I think Tom's Hardware should also include testing on Windows xp, as it still has a large amount of users. I also think they should throw in Internet Explorer 6 for the kicks and giggles.[/citation]

I don't know why you were rated down. Studies show that a very large percentage (half?) of surfers are still using XP. Your opinion was valid.

 

[przemoli]

[citation][nom]jonnydough[/nom]I don't know why you were rated down. Studies show that a very large percentage (half?) of surfers are still using XP. Your opinion was valid.[/citation]

I think he got them for IE6. I'm fine with WinXP (ON NON MAC HARDWARE), but mentioning IE6 earned another minus from me.

 

[przemoli]

For testing web browsers Ubuntu is quite good:

* Everyone try to provide drivers for it (very important on Nvidia and AMD)
* Everyone try to provide .deb's for it
* FF is updated when new version come out.

But other mainstream distros also would do (Fedora, OpenSuSE, etc.)

However you need to make sure you use proprietary drivers for Nvidia, and AMD. (and on AMD you still may not have hwd acceleration).

 

[guyhickok71]

Direct Search Frame.create TBO_IE errors are from ie8. Switch to Firefox. Chrome, anything else. I'm running Firefox now with no problems.
mayne92, your showing your age with Beta, lol I remember Beta machines and tapes. First remote in house, dang thing had a 25' cord.

 

[guyhickok71]

[citation][nom]przemoli[/nom]I think he got them for IE6. I'm fine with WinXP (ON NON MAC HARDWARE), but mentioning IE6 earned another minus from me.[/citation]

Go to the very bottom of Forums page, ther you will see a box, usually has Windows 7(most current), you can change it to XP. There are millions of threads for xp.

 

[Chef_Boyardee]

As of right now, I'm switching fro chrome to FF for a minute. Two wins in a row deserves another shot. I was a FF fan from the start, but I had switched to chrome for a couple of years.

 

[Guest]

Uninstall Firefox, defragment the disks and install the latest version again. Can you see the difference? On XP and 7 https://www.facebook.com/browser.wars

 

[amuffin]

should I feel bad that I use ie?

 

[tolham]

i alternate between chrome and firefox. i use chrome for general web browsing and firefox for porn browsing. both browsers have equal performance as far as i can perceive. chrome's advantage is tying google's services together. firefox's advantage is superior ad blocking.

 

[tolham]

[citation][nom]amuffin[/nom]should I feel bad that I use ie?[/citation]
that depends, do you use it by choice or necessity?

 

[mitch074]

@LaloFG: I never said Firefox's CSS support was bad; if you're taking it from the Maze Solver benchmark, it is a known problem: the test hits one specific use case where Firefox will do a LOT of DOM operations and refreshes for nothing - the problem has been detailed, and several fixes are actually ready, but they're all integrated in larger refactorings (especially, how to handle reflow on an absolutely positioned element if it hasn't changed in its dimensions). The day it lands, Firefox will suddenly get comparable results to other browsers, if not go faster.

@przemoli: hardware acceleration is very different on Linux, as all drivers don't support the same acceleration architectures - for example, Firefox 3.6 on Linux, using XRENDER (3.6 wasn't hardware accelerated on Windows) already had results comparable to Direct2D-accelerated IE9 on Windows on Psychedelic Wheel.

However, the new graphics stack used in Firefox since 4.0 doesn't seem to work with XRENDER anymore (at least not on my Radeon 4850 using recent AMD drivers - I should try with recent Free drivers) and Mozilla disabled hardware acceleration on all drivers - right now, WebGL alone is accelerated, and that is only with Nvidia and AMD drivers. When I try to force-enable layers acceleration, Firefox does so many rendering errors it's unusable.

Let's not forget either that there's no version of Safari on Linux, but there ARE several versions of Firefox with wildly varying results:
- the 32-bit official binary from Mozilla, that nobody installs because it's a pain (tarball only)
- 32-bit and 64-bit builds provided by each distribution, which may be outdated as Adam pointed out
- 32-bit and 64-bit builds provided by third parties: the easiest way to install the latest Firefox version on Ubuntu 10.04 LTS (which comes with 3.6 by default) is to use a PPA...

In Chrome's and Opera's cases, it's easier: both binaries are provided by their makers as a .deb or a .rpm package in both 32- and 64-bit flavours, so it is easier to test.

Of course, one solution could be to choose a platform: latest Ubuntu, in 32-bit (netbooks) _and_ 64-bit (desktop systems, recent laptops), with Nvidia graphics, using the latest versions provided for the platform, either by the editor (Chrome, Opera) or the distribution (Firefox).

 

[mayankleoboy1]

my FF8 got auto updated to FF9 only 2 days back on ubuntu.
these distros do take a lot of time to get the latest official builds.