Web Browser Grand Prix VI: Firefox 6, Chrome 13, Mac OS X Lion

[Guest]

The Mac OSX final scores do not add up.

 

[zybch]

[citation][nom]Cowboy X[/nom]The Mac OSX final scores do not add up.[/citation]
Theres a LOT about mac that doesn't add up. Just get used to it on this site.

 

[QEFX]

And the real winner is ... The end user

No more IE (which at the time sucked) vs everyone else (which wasn't supported by half the sites or your company). Now they're all good.

As a side comment: Come on Opera. You were the best ... get that HTML5 & WebGL up and running already.

 

[adriangb]

Hey tom's, nice move on the hackintosh, but you would have been much better off (from a legal stance I guess) if you used a Mac Pro and a non Apple replica (same CPU, same amount/speed of RAM, same video card, etc… heck, you might even be able to use the same board, they used to be Apple versions of Intel's Skultrail -I think).

You should also review functionality: Google's -and Opera's, to a lesser extent- native browser syncing feature rocks! You can even see Chromes bookmarks from your Google account.

 

[MrBig55]

It always makes me laugh when I see page load times. Firefox as the best add-on list, and when using the greatest ones, no adds slows it down. Of course, by default chrome is way faster than IE9 or firefox, but with adblock plus and noscript add-ons, my firefox is almost ten times faster than IE9 or chrome. It even is faster than chrome using the same add-ons on both (noscript + adblock plus)

Also I found out that chrome doesn't load web pages correctly about half of the time, IE9 has no such flaws, and it happens once in hundred of page loads in firefox. But considering how fast firefox can be with only 2 great add-ons, it's not a problem to reload a page once in a while since it saved me so much time meanwhile.

Furthermore, sometimes chrome doesn't want to load facebook games at all, or loads it but the page has so many errors that it's almost impossible to use it with these said games. Try to play Zynga's Texas Hold'em Poker with chrome you'll see what I mean: you won't ever use chrome to play to that game again even tho it loads it a bit faster (default chrome vs default firefox of course

Tomshardware.com loads in a few seconds only with my firefox 6 so there is no comparison there between any other browser. It is also a lot more secure than IE8-9 crap or the buggy chrome 11-12-13.

What is the point of testing for page load times when no other browser than firefox or chrome, got so many great addons to make web pages loads many times faster than the competition?

Also those page load times you got are ridiculous. I tried a few days ago all those: chrome 11-12-13, IE9, firefox 3-4-5-6 then my customized firefox 3-4-5-6 and my customized chrome 11-12-13. All versions of the customized firefox came way ahead of the others. Custom firefox 5 and 6 were being tied for the performance crown. I ran all these on a fully updated Windows 7 home premium 64bit.

The only firefox drawback is that when testing browser start time with prior 40 opened tabs, firefox used ALOT of memory, the same being true after I closed them all but one: it did hold on to system memory, while chrome did free most of system RAM.

So it appears that firefox is a RAM hog, and it never changed from versions 3 to 6. Other than that, it's been years that firefox is the faster of the competition, when using the best add-ons available.

 

[mindbreaker]

How about comparing ease of transferring your bookmarks from each of the other browsers. Why would I change browsers if I have to transfer 1,000+ bookmarks by hand? That's nuts. And a security test would be nice.

I am pretty much committed to Firefox because of all the bookmarks and familiarity with the features. Oh, and very big is the personalization options.

The 8 Planets add-on looks really nice: https://addons.mozilla.org/en-US/firefox/addon/the-8-planets/

"Planets" theme in gmail also looks very good with it.

Really the speed is not much of a problem as I open 53 tabs to start and then go through them one at a time. I start with my personalized Google news and just press each interesting story with the center mouse key and it opens in another tab while keeping me where I am. I can do the same with each of my tabs. Speed does not matter because the tab is loading in the background effectively and is ready when I get to it. Sometimes a couple do not load but refresh and it is there. I work left to right on the tabs and close them as I finish. Some I may leave and get back to. I probably have on average 60 tabs open at any moment.

My only complaint is that Gmail takes forever to load at sign-in and it would be nice if it loaded the first couple tabs faster than the others, so there is no delay at all. And a more fluid scrolling would be nice...I think there is a plug-in I just have not gotten to it. The memory load can get heavy. Now that I think about it I can come up with more complaints/suggestions: I am annoyed by sites that redirect immediately so that back does not achieve anything. It should be smart enough to detect that I have tried to backup twice quickly in succession and make it leapfrog that redirect the next time. I don't want to squint at a long list and try to guess which page I want to go back to. I have some problems with text being cut off. I don't want to go to the HTML just to see what it says. It did not used to have that problem. 24 inch monitors each at 1920x1200 may be part of it because the font must be large enough to see. No way I am lowering my res. I can make things larger for some sites by Ctrl + +, but video that is not maximized can still be small and hard to see and does not always have a maximize function or only maximizes to a quarter of the screen leaving the rest black. Being able to grab corners of pictures and re-size them would be nice. Being able to substitute one color for another would be nice...white can be a pain. There are some options there but no substitute. Oh, one more thing, If I force a shutdown with windows task manager or it crashes I want every time to be able to select which windows and tabs I want it to restore. Sometimes I am given the option and sometimes I am not...not sure what is up with that.

I have tried Chrome but I don’t see the point heck I can’t even find any buttons to push. Probably just don’t have the patience.
Explorer is just boring no appearance customization to speak of. And I have no idea how to open a bunch of tabs simultaneously. Don’t know how to do that in Chrome either. In Firefox it is a snap.
I know I am not impressing anyone with my lack of expertise but I am happy. The only browser I am interested in is Firefox. All I can do is hope they fix the things that bother me.

I would like more of a tree approach in the forward/back functions. For example if I pres forward and I went to two different places don't just assume the last one open both in a split screen or popup a choice with small versions of each to choose from. Recovering closed tabs would be nice too. It can just show you a list of windows/tabs you closed starting with the most recent but with miniature versions say 1/32 sized. And you can just scroll down. In privacy They could have an option to not use that feature, fore worried people. But I sometime close something I did not intend to for one reason or another and I can't always remember how I got there especially if it involves some Google search I misspelled or something.
Oh, and one last thing I hate it when I enter a comment somewhere where I was not signed in and when signing in it kills the comment. I don't want to rewrite that all the time. And I can't always remember to copy and paste to Word or just copy.

 

[adriangb]

I read some more comments: I take back what I said. You don't need a Mac Pro + non Apple PC, you need a Mac + Bootcamp. So you get 100% native OS X and 100% native Windows. As for the hardware, it's true that a lower end system would be better, but remember that it MUST support 64b and modern instruction sets. I think that the ideal would be the oldest Core 2 Duo mac that you can get your hands on that supports Lion. My brother has a 2009 MacBook that has a Core 2 Duo + 9400GT, so it supports video acceleration (?). This would be an issue chosing the system, it would have to support GPU hardware acceleration.

Also, since you already got into hacks, could you build a hackintosh identical to a Mac Pro and install OS X on it fully (I'm sure some guys on forums can help iron out all the details), and then post extensive benchmarks comparing the performance? Would be fun Also comparing what 2000 can get you in mac hardware vs PC hardware...

 

[rockford1950]

I agree with MrBig55 regarding load times of tomshardware.com with Firefox and Chrome. tomshardware.com (and most other sites) load faster, too fast to time, with Firefox 6 with AdBlock installed vs Chrome 13 with Adblock. (MacOS X 10.7.1). And the testing overall misses the mark on "what" to test. Yes, HTML5, Javascript and the like are important to test. But in the old days, when bandwidth was limited, we used to test how fast certain graphics were rendered. Certain browsers handled GIFs faster than JPEGs or PNGs. We could skew tests in favor of one browser over another by choosing test sites that had the preferred" graphic types on it. I expect that's still the case with today's browsers. So to do fair tests, you need to choose sites with equal amounts of graphic types.

And if you care at all about business environments, you need to test authentication support. Opera doesn't support Kerberos and Safari & Chrome have problems with Kerberos authentication on IIS servers, so you're only choices for single sign on Kerberos authentication in the workplace are IE and Firefox (tweaked for SSO.)

 

[Guest]

I'm sorry, I don't believe the results. I work with dynamic Web and is very skeptical about IE JS performance results. They are either paid for or wrong, there is no third option. IE sucks at JS performance and DOM, be it IE8, IE9 or IE10. And there all tests show just unbelievable JS performance values for this browser.

 

[adamovera]

[citation][nom]Cowboy X[/nom]The Mac OSX final scores do not add up.[/citation]
[citation][nom]zybch[/nom]Theres a LOT about mac that doesn't add up. Just get used to it on this site.[/citation]
Yes they do, don't forget any ties.

 

[adamovera]

[citation][nom]Asmod4n[/nom]Safari supports WebGL, you have to enable it in the Developer Menu.[/citation]
Yeah, when Apple wants to enable it by default, we'll test it. The same could have been said for Chrome awhile ago, Google eventually enabled it and that's when we tested it.

 

[adamovera]

[citation][nom]adriangb[/nom]Hey tom's, nice move on the hackintosh, but you would have been much better off (from a legal stance I guess) if you used a Mac Pro and a non Apple replica (same CPU, same amount/speed of RAM, same video card, etc… heck, you might even be able to use the same board, they used to be Apple versions of Intel's Skultrail -I think).You should also review functionality: Google's -and Opera's, to a lesser extent- native browser syncing feature rocks! You can even see Chromes bookmarks from your Google account.[/citation]
Violating a TOS isn't a crime (yet, anyway), just a warranty voiding action. Like jail breaking your smart phone: Sure, go ahead. Just don't expect any help if you screw it up. Selling systems pre-installed with OS X on the other hand is a sue-able endeavour, remember Psystar?
Like I said earlier, this crowd is all about building custom systems, so using a genuine Apple with Windows is less likely to appeal to the core TH crowd than OS X on a custom PC. But Apple is welcome send us any model of Mac they wish to test at any time for a rematch.
We compared the syncing services of Chrome, Firefox, and Opera in WBGP4 - http://www.tomshardware.com/reviews/firefox-4-internet-explorer-9-chrome-10,2909-3.html

 

[adriangb]

[citation][nom]adamovera[/nom]Violating a TOS isn't a crime (yet, anyway), just a warranty voiding action. Like jail breaking your smart phone: Sure, go ahead. Just don't expect any help if you screw it up. Selling systems pre-installed with OS X on the other hand is a sue-able endeavour, remember Psystar?Like I said earlier, this crowd is all about building custom systems, so using a genuine Apple with Windows is less likely to appeal to the core TH crowd than OS X on a custom PC. But Apple is welcome send us any model of Mac they wish to test at any time for a rematch.We compared the syncing services of Chrome, Firefox, and Opera in WBGP4 - http://www.tomshardware.com/review [...] 909-3.html[/citation]

Yeah, I know it's not a crime (I own a hackintosh myself…) -it's just not very well seen by some.
And -even thou a mac with bootcamp might not be that appealing- it would be a better test: both OS's run nativley (even thou you can make a near perfect hackintosh, the're always some things that are workarounds, or just outright don't work.

It's great you're repling to the coments

 

[adamovera]

[citation][nom]AlexAT[/nom]I'm sorry, I don't believe the results. I work with dynamic Web and is very skeptical about IE JS performance results. They are either paid for or wrong, there is no third option. IE sucks at JS performance and DOM, be it IE8, IE9 or IE10. And there all tests show just unbelievable JS performance values for this browser.[/citation]
In the four combined JS/DOM tests IE9 gets two 4th place finishes and one last place. The only 1st is in SunSpider, which is quarantined.

 

[fstrthnu]

So Firefox on Mac sucks compared to Windows. That's not very reassuring...

Now I'm VERY curious as to how much Mozilla improved the performance for Firefox 7. I can't really give accurate results because I usually run with LOTS of tabs open (don't ask why), but it does seem a little peppier on the Firefox 7 Beta than Firefox 6 stable before.

 

[bmouring]

Whaaa? No lynx? Just because it can't run most of those benchmarks... feh, philistines

 

[Guest]

I LOL'd at the "Java and Silverlight" comment.

 

[sundragon]

[citation][nom]adriangb[/nom]Yeah, I know it's not a crime (I own a hackintosh myself…) -it's just not very well seen by some.And -even thou a mac with bootcamp might not be that appealing- it would be a better test: both OS's run nativley (even thou you can make a near perfect hackintosh, the're always some things that are workarounds, or just outright don't work.It's great you're repling to the coments[/citation]

1. From what I understand you have to purchase computers for testing purposes. Plus I see so much negative press on this site for Apple, I wouldn't be holding my breath, lol.
2. I have to agree with adriangb - You should test on a Mac natively as hackintoshing usually has subtle incompatibilities in drivers which can skew results. Bootcamping Windows 7 and then comparing so both OSs run natively.
3. Glad you are responding to the comments

 

[harsh86]

Peacekeeper is a really bad benchmark. Do some research and you'll see that it needs to be in quarantine.

One out of the many reasons its pretty bad is its over use of the date function in its testing harness which can often equates to being a 40-50% of a performance hit on some of its micro-benchmarks. Some Browsers are really fast at calling date but some aren't. This would be fine if the point of the benchmarks were to test how fast the browser can create a Date, but it's not. In fact creating a date millions of times a second is only really useful for looking good on benchmarks, uh um, chrome. This is because dozens of bad benchmarks share this problem. A benchmark with a harness that runs significantly faster in some browsers and slower in others is an automatic fail. The harness is the referee and thus should be designed to be impartial. Its called a fair test, which most people learn how to carry out in high school science class!

 

[oneblackened]

Firefox, for me, has always been more reliable. I also am not a fan of chrome's GUI. Far too annoying with a lot of tabs open because they don't scroll, they just get too small to remember what is what.

 

[retrig]

Chrome still doesn't have NoScript or AdblockPlus. Worthless until then.

 

[kittle]

you definitely need to measure useability of the browsers.

I prefer to use Opera and/or chrome. but here at the office both of those fall over when presented with a corporate firewall. Chrome tries (and fails badly) to use your existing settings, and Opera cant seem to figure out how to re-authenticate when my proxy setting has expired, forcing me to restart the whole browser.

Firefox has no issues with our corporate firewall.

 

[mchuf]

[citation][nom]cookoy[/nom]And to think Apple hates Flash...[/citation]

Thanks to CS5, Apple hates Adobe, which also owns Flash. Thus Apple hates Flash.

 

[mindbreaker]

I took a look at Chrome again. It was easy to import my bookmarks, so I have to give it credit there. However, some images were broken when I opened my 50+ usual tabs together. It also would not close quickly it took like 6 seconds to close. I also could not see any real controls...they must be hidden somewhere. Scrolling seemed better..it felt very responsive. It also was using a lot of memory. It did not look like much but it had more than a dozen processes so you have to add them up.

I also went to the add-ons at Firefox...there is just too much good stuff to abandon Firefox! There are more than a quarter million different personas! Doubtless, there is something you will like. I got about a dozen more. I am using one called A Blue Planet...looks nice. Actually, I downloaded 26...and I am very picky.

And of course there are all kinds of other add-ons dealing with function. They even have scrips. Can't see the point though as I can open the tabs I want with one click and having it browse without me is pointless. And if I want to download a complete site there are other ways to do that. What does that leave? Well, someone may have a use.

One more thing Firefox should add is, that when a link is clicked and it opens in another tab and you click "back" it should go to the page that opened it rather than doing nothing...should go to that previous tab.

 

[Guest]

Installed Chrome the other day. I logged in as Administrator and installed it: it installed only for that user and for no others. Didn't even ask me if I wanted to install for all users. Anyone else get that? Completely useless!