FireFox heading for 15% market share?
blogs.lessthandot.com
blogs.lessthandot.com
It's the only browser which isn't a corporate puppet, respects the user's requests and has a truly open development model.
It's the only real browser if you ask me.
So, Chrome is to Firefox what Firefox was to Internet Explorer?
Firefox does get most of its funding from Google currently, but there are alternatives. I'm pretty sure Microsoft would also be happy to fund Firefox in return for putting Bing by default there.
> Try konqueror for something more like what you're describing.
onqueror uses WebKit, and Google is the main contributor to WebKit, so you aren't getting away from Google that way.
Lately, Chrome has begun to ask me to "log in to Chrome". What? No thanks, I don't trust Google never to associate my browsing habits with my Google account. They have too many incentives to go evil.
If RMS can survive on a 9" netbook for the sake of his principles, a lot of us can survive on Firefox. Besides, it's not as terrible as Chrome & Safari aficionados will have you believe. There is definitely room for improvement, but none of those are deal breakers for me personally.
Then there are bugs that have been around for years now, like pressing Cmd+Left or Cmd+Right on a textarea does nothing. (Windows users: These are the OS X equivalents of Home and End) They say that there's a conflict with another set of key bindings, and that this is an issue with XUL, but I strongly believe that the Firefox UX should dictate how XUL works, and not the contrary.
The last time I looked, they were still struggling to make each tab an independent process. Chrome first shipped more than three years ago with such a feature. Since then, both Safari and Internet Explorer have adopted a similar model.
Their new, rapid release cycle is a great improvement. But updating Firefox is still annoying and slow. You have to check for extension compatibility. You have to wait for a progress bar to fill. They should have nailed this when 4.0 first shipped.
I wonder if Mozilla is suffering from finantial/funding issues right now, already, since their main source of revenue, ads on Google results in Firefox, is constantly decreasing. How long can they keep mantaining Firefox? Should they switch to a crowdfunding model?
It's really a shame this is hapenning, since they have great intentions: To keep the Web open and innovative. In terms of philosophy, they are also on a much higher ground than Apple, Microsoft and Google.
It sounds like your priorities - UI features like scrollbars - are not identical to the FF developers. That's fine - it's why it's good we have several good browsers to choose from.
Generally though, Firefox has definitely been inventing and adopting plenty of new and exciting technologies, some random examples: TypeInference in the JS engine (still unique to Firefox), Layers graphics system with support for hardware acceleration/Direct2D/etc., WebGL (together with Chrome, lots before Safari or Opera), Audio Data API, new JS features like Proxies and WeakMaps (which Chrome yesterday blogged about finally implementing), etc.
> But updating Firefox is still annoying and slow. You have to check for extension compatibility. You have to wait for a progress bar to fill. They should have nailed this when 4.0 first shipped.
FF 10 defaults addons to compatible, and FF 11 and 12 will add things like update in the background (no progress bar, no prompt). It would have been good to have had this with 4.0 of course, but it does take time to do - waiting to release 4.0 until now would have meant not shipping all the other great features mentioned before.
Firefox is the only browser shipping a fully standards-compliant implementation of IndexedDB. IndexedDB is one of the most important features of HTML5. Clearly Mozilla is moving fast enough, their priorities are simply different than the grandparents.
Mozilla has had their priorities wrong for ages. They still coast on the glory days of the 2000s when they were the good guys, even though they're now just another browser maker with millions in the bank and a giant legacy platform to take care of.
Bugzilla is full of code-centric developers who seem more interested in theoretical feature sets than actually making features that are useful to end-users.
Take for example the differing approach to sub pixel rounding. WebKit's approach recognizes that rounding errors cannot be viewed in isolation, but can result in visually disruptive artifacts in group. Mozilla still prefers mathematical purity over reducing disruptive artifacts, with e.g. text drifting perceptibly against a container when it is being moved slowly.
But the best example was border-radius: Moz implemented it for years, but didn't anti-alias rounded corners. There wasn't even a guarantee that identical corners on the same box would be rendered the same, due to rounding errors. When WebKit implemented it, they had anti-aliasing from day one, and suddenly everyone was using it. Big surprise.
The web has always been a visual medium, and the stick of platform APIs needs to be sold with the carrot of attractive design and seamless rendering. In their marketing materials, Moz gets it, but in their development strategies, they are still way behind.
About ACID3, the main issue was SVG fonts. They are an unneeded standard, because Web Fonts exist, and Firefox has therefore not implemented them (so saying they eventually did is incorrect). The right thing to do was to fix the ACID3 test, which has been done, see more details here:
http://limi.net/articles/firefox-acid3
About 3D CSS, sure, there were disagreements and eventually Apple got it's way, that's a normal part of the standards process, you can't predict how things will work out. You can't expect every browser maker to be right from the beginning on every topic.
For example this post, which has been removed by the author but is still referenced elsewhere: http://blog.mozilla.com/rob-sayre/2008/03/26/acid3-is-basica...
The gist of it was that ACID3 didn't test things that were relevant to 'real' web developers. Which is a dubious criticism at best, seeing as ACID3 was an excellent weathervane for sorting the wheat from the chaff.
But it's hypocrisy in hindsight given that all the great features that web developers flocked to came from the other engines. And now Mozilla is complaining that there are -webkit prefixes everywhere.
As for 3D CSS, it's not just about 'disagreements' or 'the spec'. The work Apple has done on iOS and desktop has made WebKit2 a seriously impressive engine. I made this site recently: http://acko.net/. It's butter smooth in Safari, choppy and ugly as hell in Firefox 10 (if I enable the 3D effect there). And yet Mozilla goes around proclaiming Firefox now supports CSS 3D and is on par with WebKit.
? It isn't Mozilla saying this. It's every browser but WebKit. It's Microsoft, Mozilla and Opera, and non-browser people too.
The mobile web is broken because it's full of WebKit-specific stuff that other browsers can't render. That's a fact. Mobile version of IE, Firefox and Opera can't render the mobile web properly because of that, even though they are standards-compliant browsers.
> As for 3D CSS, it's not just about 'disagreements' or 'the spec'. The work Apple has done on iOS and desktop has made WebKit2 a seriously impressive engine. I made this site recently: http://acko.net/. It's butter smooth in Safari, choppy and ugly as hell in Firefox 10 (if I enable the 3D effect there). And yet Mozilla goes around proclaiming Firefox now supports CSS 3D and is on par with WebKit.
Link? All I've seen is that Firefox was announced to support 3D CSS stuff, which it does. There might be bugs where it doesn't render as fast as it should, you might want to file one about your website (which is cool, btw).
Side note: WebKit2 is a new embedding API for WebKit that is multiprocess-enabled, not a new version of WebKit. Chrome, for example, doesn't use WebKit2 - Chrome has its own multiprocess code. The 3D CSS stuff is in WebKit, and not related to the WebKit2 API (which, btw, has some performance problems currently that Safari is suffering from, but not Chrome).
As for the mobile web... even my WebKit Android browser fails often, even on this very text field. There is no single 'webkit for mobile', and it's all the non-iPhone users that are treated as second class citizens, webkit prefixes or not. That's IMO a very different problem.
As for WebKit2, you're right, that was badly worded. What I meant was "the Safari engine since they switched to WebKit2" (i.e. Lion). It came with a noticeable improvement in rendering performance, which Chrome has not been able to match on any system I've tried.
Ultimately it all comes down to opinion and perception. But I have a hard time finding people who are still excited about Firefox who aren't Mozillians.
It's the same problem, and it's just as bad: If there is one dominant implementation, and other browsers can't render the actual web even when they implement the relevant standards - then the web is broken. If it's WebKit vs the rest or iOS vs the rest, it's just as bad.
Regarding your frustration on the Mozilla bug tracker, I've definitely felt that both there and basically on every other major bug tracker. The other day I filed a bug on Node.js, and in 15 minutes there were three responses, saying things like "why is this a bug in node?", "don't file bugs on node like this", and soon after the bug was closed. A few days later, it was confirmed as valid... that's how it goes in big bug trackers ;)
> Ultimately it all comes down to opinion and perception. But I have a hard time finding people who are still excited about Firefox who aren't Mozillians.
I have very much the opposite impression, we probably just visit different websites and communities.
http://www.0xdeadbeef.com/weblog/2010/06/intellectual-honest...
Two years later, Firefox 10 comes out, the first version that actually supports those effects.
Separately, the standards process worked itself out, and then Firefox implemented the standard. That sounds ok to me.
> The last time I looked, they were still struggling to
> make each tab an independent process. Chrome first
> shipped more than three years ago with such a feature.
> Since then, both Safari and Internet Explorer have
> adopted a similar model.
Gecko is old, and I don't envy the task of the Mozilla developers who are responsible for balancing the advantage of modern enhancements with the risk of breaking addons that have been working for the past decade or more. From this perspective, it's no wonder they're looking towards writing a brand-new browser engine.
All browser engines are old. WebKit's code begin in 1998, for example, and IE and Opera also have plenty of code from the 90's.
They are all huge C++ codebases too - there are big burdens for all current browsers that stem from the code's age and size.
Gecko does have the additional task of keeping addons working (which WebKit doesn't), but on the other hand WebKit has the additional task of maintaining a stable embedding API (which Gecko doesn't). Perhaps IE and Opera don't have those burdens, they are closed source so we can't tell.
> I don't envy the task of the Mozilla developers
For the reasons mentioned above, you can say the same of all the other browser developers. Or, also for all OS developers, even a relatively recent OS like Linux is quite old and has all manner of baggage in terms of maintaining compatibility with legacy kernel modules and such.
Big, long-term projects like Gecko, WebKit, Linux, Windows, etc. are hard!
As long as we are comparing browser engines, though, I'd like to stress the relative scarcity of resources available to the Firefox developers. Mozilla's recent windfall with respect to their arrangement with Google is great, but Mozilla's attention is focused on much more than Firefox nowadays--Boot2Gecko, BrowserID, and Rust/Servo being their most prominent diversions at the moment.
Today - this wasn't true until recently - I think that Mozilla does have the resources to push additional projects than Firefox.
I wish they would prioritise their UI before they do this, though. Mozilla has one of the most talented UI-teams in the browser landscape based on the mockups I've seen. Take http://blog.stephenhorlander.com for example. That looks absolutely amazing, yet I don't believe more than 20% will make it into the final product, since this has been the case before.
If Mozilla is prioritizing other things as you mention in another post, it's time to flip those priorities around for a while until they've got what their mockups are showing. Down to the smallest detail.
Everyone pushes features galore these days to the point no-one knows what "HTML" is objectively defined as. People do know what subset works consistently across all browsers though, which Firefox supports reliably.
As for Firefox's UI - they maintain consistency between platforms. Apple like to throw absolutely horrid curve balls every OS release and change just about everything for the sake of some misfeature they decided to make their users suffer (scrollbars in Lion being one of them). Also TBH compared to the ABI stability of Windows and Linux, OSX is a pile of shit to develop for so things are going to take time if you have to sync a lot of code across a lot of platforms. If they didn't bother - so what? At least your browser world didn't cave in due an OS release.
Regarding the whinge about scroll bars. I'm sitting on a Windows 7 machine at the moment and the scrollbars work the same as they did 16 years ago i.e. how I expect them to work and consistently with everything else on the machine. New scrollbars is not a feature - it's an annoyance.
It's called innovation, and I'm glad it's still happening.
It's not an improvement.
So I switched to Chrome and got used to their Inspector. I missed padding and margin being shaded in different colors, but then I didn't have to restart the browser after every 5 page loads when doing heavy JS UI development.
The newer builds of Firefox are fantastic. It loads instantly and the JS engine has been patched up. But it's too late.
Firefox still runs poorly for me and until that changes they have no hope of winning me back. Who knows, maybe it's partly my fault that it doesn't run well or maybe there are some settings I can tweak to boost it's performance but even if that were the case I want my browser to work out of the box.
Funnily enough, I have FF on my older laptop (dual core 1.2 gHz, 1.5 GB RAM) for blocking a bunch of useless scripts that slow me down and drain my battery.
I think Firefox power-users will stay because of its flexibility and advanced options and all those things they like but normal users don't care about. But if you look at it from an average, every day consumer's point of view, Firefox fails to provide an experience as smooth and nice as Chrome's.
Stability wise, Firefox is way behind. At least in my situation. I have a custom PC, Windows 7, no bloatware, extremely stable, my computer is ON for months without being restarted. I've got Chrome up and running for weeks and weeks at a time without it being shut down and reopened with over 80+ tabs open at all times. I kid you not. 80+ tabs. I keep them up because if I file them away as bookmarks I'll never see them again, so they just kind of stick around. I can still photoshop, and illustrate without having to close Chrome. When I try that with Firefox the poor browser falls flat on its face. After a few hours of having all that open my system starts to get sluggish. Most of the time the Firefox crashes due to an unresponsive script or flash. In Chrome only 1 window crashes and I can just refresh it, in Firefox the whole browser crashes. -_-
So little by little I migrated to Chrome, so did the rest of the family. It's easier to work with, less dialog boxes, silent updates, extensions don't break when you upgrade, etc...
MORAL OF THE STORY: UI & UX matter, a lot.
It crashes all the time with like 30 tabs open...freezes up...and it uses 900mb of ram just by simply browsing....and can spike up to 2.3-2.4 GB ram if you do visit any intensive site.
The only reason I don't switch to Chrome, is because it doesn't support a lot of plugins that I use.
Sorry to see you get downvoted by the Firefox fanboys.
1) The Awesome Bar is really just that. I have not found any other browser that works as well in this area as does Firefox. 2) Firebug is one of those tools I've been using so long that it just seems like part of the browser. Safari's Inspector is pretty dang close but still not what I'm accustom to.
I'd love to switch to something lighter. I just don't see that happening as a permanent switch anytime soon.
In my opinion it would have been better to wait with things like the rapid release cycle until now, but well... at least that's fixed now.
Also I use Aurora (Firefox 12) and it feels really nice. Quick, nice features not feeling bloaty at all. If that direction remains I am sure many users switching to Chrome could come back. It's also doing pretty well when it comes to web standards. The only problem I still see on the web is that developers sense for UserAgent strings instead of features, which is just plain stupid and incompetent.
As the mobile web explodes, and Mozilla is barely even competing in that market, then their overall market share will fall.
The new Firefox Mobile, currently in beta, is quite good - it uses a native Android UI and is very fast and smooth. The older version, which used XUL, was indeed sluggish.
But, to be fair, none of this really matters. iOS and Android are by far the top mobile OSes, and hardly anyone replaces the default browser on them, which makes it extremely hard for alternative browsers (Firefox, Opera) to make any headway. Tech people like HN readers might, but 95% of people just use the default browser on their phone.
The performance cost of XUL, though, is just too much for mobile.
What what it's worth, I've long stopped using Firefox since the release of Chrome. The constant dealing with extensions not even trying to run after an upgrade and the need to constantly change settings in about:config to get extensions to function has caused me to abandon ship.
It also seems like Mozilla is frantically trying to be more and more like Google Chrome with the rapid release cycles and plethora of other features ported over.
I sorta like Chrome, but I have found Tree Style Tabs for Firefox to be indispensable so I'll stick with it. Firefox's extension method is both very powerful (gives access to browser internals) and very brittle (gives access to browser internals). That, to me, means more-useful extensions that might break, which I'm okay with.
I am using Aurora now and Netbeans and my IM client are using more memory than Firefox. Many of the problems from the past are gone now.
Anyway, I can see how someone could choose to stop using Firefox because of the new version number policy. Especially if those people are switching to Chrome :)