Firefox 6: A Firefox too far? (Review)
Summary: Firefox 6 is better than the Firefox 5 Web browser. That's the good news. As for the rest, well I'm not impressed.
I used to love Firefox. For many years, it was the best Web browser around. Then, it got old and cranky. The good news is that Mozilla got serious about improving it. The bad news is that everyone else, even Microsoft with Internet Explorer 8 in 2009, did too. The other Web browser developers have caught up and surpassed Firefox. Worse still, the last few versions of Firefox haven't been that good. Firefox 6, the newest version, is better than Firefox 5, but it's not that great either.
Like most modern Web browsers, Firefox 6 has a "less is more" style interface. If that's not to your taste, and I confess I like having real menus, Firefox does make it easy to bring back a more traditional Web browser interface.
Firefox also now supports, ala Google Chrome Web browser, pinned tabs. With these I can always have my favorite Web sites ready to go in tabs kept pinned to the left side of the tab bar.
I also liked that Firefox is continuing to work on improving its security. Firefox is about as safe as a browser can get these days.
This version of Firefox also does much better than previous versions at managing memory. Unfortunately, you really need to be a techie if you want to use the about:memory option to bring Firefox's endless hunger for RAM under control. While I'm a techie myself, and I appreciate this level of access to the program's memory use, I think 99% of users would have been happier with even better, and invisible, memory management.
I know that Firefox 7 will have better memory management. I also know that Firefox 7 is due out on September 27th. I would have been just as happy to wait for that and then for Mozilla to have released Firefox 6. As it is, Firefox 6 feels like a placeholder.
Don't get me wrong. Firefox 6 is better than Firefox 5. That last version of Firefox crashed more often than a drunken teenager behind the wheel of a sports car.
OK, that was the good news. Here's the bad news.
First, I, and a lot of other people, am already sick and tired of Firefox's manic major release schedule Come on! Stop spending to much time playing with your major release designations, and just focus getting the software right, OK?
This isn't just me being grumpy. I've found one Web site, my bank as it happens, that works fine with Firefox, but it doesn't recognize Firefox 6 as being a valid Web browser. Yes, some of that is sloppy programming on their side, but Web browsers have to be ready for messy Web sites. As it is, I "fixed" the problem by using User Agent Switcher to mask Firefox 6's Web browser identity with an older version of Firefox's identity.
Firefox 6: A first look (Gallery)
At least that Firefox extension would work for me. Far, far too many others don't work. Mozilla developer, Dave Townsend, claimed earlier this year that "Faster releases are good for add-ons." No, they're not.
Just because Mozilla developers are paid to keep up with constant change doesn't mean that individual developers or small independent software vendors (ISV)s can keep up. From what I'm seeing they're falling behind supporting Firefox with its six-week release schedule.
Even the big ISVs seem to be having trouble. For example, Adobe Flash on Firefox 6 on Mint Linux is failing for me. Chrome's built-in version of Flash works just fine on the same Web sites on the same version of Linux. Go figure.
I also have to say that when it comes to raw performance, Firefox isn't blazing into the lead. Far from it.
I've been using Firefox 6.0 on both my Linux and Windows XP and 7 PCs. For benchmarking, I used my Gateway DX4710 running Windows 7 SP1. This PC is powered by a 2.5-GHz Intel Core 2 Quad processor and has 6GBs of RAM and an Intel GMA (Graphics Media Accelerator) 3100 for graphics. It's hooked to the Internet via a Netgear Gigabit Ethernet switch, which, in turn, is hooked up to a 60Mbps (Megabit per second) cable Internet connection.
On this Windows system, I ran not only Firefox, but Chrome 13 and Internet Explorer (IE) 9 to see how Firefox 6 compared to its chief rivals
Firefox, once more, didn't do well on Web standards. On the Acid 3 compatibility test, which checks out how well a browser complies with various Web standards such as CSS, JavaScript, and Extensible Markup Language (XML), Firefox had a score of 97. Chrome had a perfect score of 100, and IE had a 95. These were the same marks for the browsers the last time I checked in on them.
On the HTML5 Test, which checks to see how compliant the Web browser is with the HTML5 Web page standard, Firefox showed a significant improvement. This time, it came in with 313 points out of a possible 400. Unfortunately for Firefox, Chrome did even better with a score of 340 out of a maximum of 400. IE? Microsoft can say what they will about HTML 5 support, but it still came in dead last with 130 points.
As for pure performance, I first tried the browsers out on Mozilla's own Kraken 1.0, which is their update of the SunSpider JavaScript benchmark In Kraken, the lower scores are best. Here, Firefox was, at best, OK with a score of 7588.2ms. That's a trifle slower than Firefox 5. Chrome 13 zipped into first place with 4927.7ms. IE? Ahem, IE came with a stinker of a score 17,051.9.
With Peacekeeper, where higher scores are better, Firefox really fell behind with 4,588 points. On this benchmark, IE surged ahead with a score of 8,343 to Chrome 13's 8,010.
According to Google's own JavaScript benchmark V8 Benchmark Suite, where higher scores are better, Firefox does a mediocre job with a showing of 3,614 compared to Chrome 12's 7,677. Firefox scored much better than IE way, with its 2,193, but on this benchmark it's not even in Chrome's ballpark.
As for SunSpider 0.9.1, the oldest JavaScript Web benchmark, where lower results are better, Firefox shows well with 301.2ms. Unfortunately, Chrome 13 with 289.3, and IE 9 with 252.6 are still better
And, when all is said and done, that's also my verdict on Firefox 6. The other browsers are better. While I can't recommend IE 9 for all Windows users, since Microsoft has seen fit to not make it available to XP users, I can recommend Chrome for everyone. I wish I coud still recommend Firefox, but I can't.
Related Stories:
Firefox 6 already sees bump in traffic soon after debut
Mozilla wants to hide Firefox version numbers from users - UPDATE
Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.
Talkback
RE: Firefox 6: A Firefox too far? (Review)
RE: Firefox 6: A Firefox too far? (Review)
Wrong.
A vast majority of addons work with Firefox 6. Stop spreading ignorance.
RE: Firefox 6: A Firefox too far? (Review)
... Only to be broken again with v7.
RE: Firefox 6: A Firefox too far? (Review)
Do you always judge by beta?
RE: Firefox 6: A Firefox too far? (Review)
Not always, but it is a sign of things to come. If I was a developer, and my addon was breaking every six weeks, I'd think twice before bothering to update it again. Developers don't always have all the time in the world to continue updating their addons just because Mozilla wants to artificially inflate their numbers.
I still am still in the process of trying to figure out how to uninstall addons and plugins from Firefox that haven't run since I left FF3.5...
RE: Firefox 6: A Firefox too far? (Review)
Soon, Mozilla will be removing the version number check, before the end of the year. Not sure if it's going to be in version 9 or 10.
RE: Firefox 6: A Firefox too far? (Review)
Sweet. Long overdue, but good news nontheless.
RE: Firefox 6: A Firefox too far? (Review)
The only time I want firefox to shut down my add-ons is if it detects that not doing so would crash the browser or something.
RE: Firefox 6: A Firefox too far? (Review)
And I have begun to use IE9 again. I would love to remain with Firefox, as I like the product and want to encourage the development of all, and any, software.
RE: Firefox 6: A Firefox too far? (Review)
Please tell, is IE9 as *functional* as FF4??
My main needs..
adblock using Fanboys list..
session manager compatible with tabmixplus.
Noscript, yescript, abilities..
flashblock with whitelist.
fully customizable, all buttons, menus moveable.
video & pic downloader like downthemall.
exeteded statusbar & download status
- has IE stopped lying about page download?? try unplugging your net, then load page! the 'progress bar' still moves!!... :(
RE: Firefox 6: A Firefox too far? (Review)
(note: there is NO 'reply' button where I want it????)
Stop spreading ignorance? That seems to be the norm around here.
RE: Firefox 6: A Firefox too far? (Review)
But anyway I'm running FF7b so some bugs and absurdities are expected. It scored faster on tests though
Total: 5995.0ms +/- 2.2% <--Kraken
Firefox(v7.0) Scored: (Peacekeeper)
4999 Points
As evidenced by these two scores, I think improvement is being made with the fast release schedule. They may not be significant improvements but the fact that you wait far less for an improvement on Firefox even if it's only a bit is nice if you ask me.
RE: Firefox 6: A Firefox too far? (Review)
Fast releases are OK and wanted for security fixes, so they are OK within a major release branch (and these fixes have then to be integrated also in other active branches).
But functionality fixes, or even speed improvements should not be posted so early if it requires a new major version for compatibility checking.
Those users that will want faster releases could subscribe to faster update channels, but this should not be the default (and even Chrome uses separate channels, with the default channel being the one with the slowest updates, except for security fixes or fixes against crashes; Chrome users that want faster releases can subscribe to the beta channel or to the alpha/development channel : I think it is an excellent idea that limits the impact of incompatibilties and can still allow early detection of those incompatibilities long before they come into the main channel)
Chrome also develops its releases with distinct milestones for future releases, all developed in parallel but where new functionalities are scheduled long in advance; there's no surprise update for developers of addons or websites.
Mozilla should be inspired by how Chrome can manage its versioning, and still being able to perform fast updates, without breaking existing functionalities or backward compatibility. The way Mozilla is managing Milestones is that this is only a large bag collecting ideas or concepts without clear specifications, and without scheduling them; so when a milestone is about to come, it just reports to the next one those that will not be implemetned immediately. The next milestone is then full of many crap ideas that may only have experimented from a past version, and no longer workable; there are many unfinished tasks in those milestones, and no work plan to resolve them by dependency or in an ordered way (in fact I think that even the developers of Mozilla don't seem to agree on any planning, each of them is working separately with too little interaction and cooperation). So most of them are just overwelmed by long and unorganized lists of bug reports or RFE's to resolve...
RE: Firefox 6: A Firefox too far? (Review)
He obviously has a strange config'
RE: Firefox 6: A Firefox too far? (Review)
He says he is using Linux Mint 10, with Firefox 6. As teh latest Linux Mint isn't up to Firefox 6 yet, he had to install it himself.
He might have gotten some of the dependencies wrong.
I don't want to start a flame war but...
Dependencies?
When has that ever happened with Firefox? Generally, if a program will run, the dependencies are there. I can't recall ever tracing program crashes, other than perhaps a couple of crashes on open, back to missing dependencies and only occasionally to out of date dependencies, and that was in Windows.
Well, Manually Tracking Dependencies Has Its Own Pitfalls
Generally something like "dll hell" doesn't happen in Linux because libraries are versioned well (and that's gotten better in Windows too, I think), and because dependencies are managed by the package manager.
However, if you start compiling your own software, or installing simple executables in Linux that rely on open source libraries, then you can start having to track dependencies for those programs yourself. That can be annoying, but I've never seen it cause random program crashes. Usually, it just means you have to track down and install additional libraries to get one of these programs to work.
Proprietary software in Linux usually contains it own dependencies (this is often true of software for Windows as well). It would be conceivable for proprietary programs to create folders for standard proprietary libraries in /opt, and for that to lead to different programs overwriting these libraries. Then you could see problems similar to "dll hell" in Linux. Problems like that, however, haven't reared their heads in Linux so far. Overwriting a dll is increasingly unusual in Windows as well. In Windows more recently, the problem you sometimes encounter seems to be that different versions of the same library exist in multiple places and trying to figure out which dll files programs are actually using (probably from the PATH variable) and making them all work.