Firefox May Already Be Dead
pcworld.com
pcworld.com
There are plenty of FireFox plugins I use (primarily FireBug, but Chrome has good debugging tools and it's only a matter of time before someone builds an equivalent of FireBug for Chrome).
AdBlock Plus and NoScript give me not only less junk but block an awful lot of attack vectors. I feel kinda vulnerable without NoScript.
Oh, and a full linux-build, please. I've tested Chrome on Windows and am impressed, but as long as there's not linux build I can't use it. OSX people probably feel the same way.
The OS X battlefield will be interesting because Safari is so prominent. If Apple/Webkit decide they want to compete with Chrome, they're in an excellent position to do so - and they already have Javascript disabling, ad blocking, and GreaseMonkey support in Safari, not to mention a browser that's incredibly fast.
IMHO, these guys need to do some more marketing.
EDIT: Sorry, I'm looking for such features in Chrome.
I usually navigate by searching for some substring of the link.
And yeah, Firemacs is awesome. As a matter of fact, it does allow for navigation using some vi bindings by default. Even though I'm an Emacs guy, I find using hjkl a lot more comfortable than C-b C-n C-p C-f. I've heard great thing about vimperator, too.
Adblock is better but privoxy helped me to change. Only open Firefox(which in comparison takes like ages) for development.
Mozilla's lagging pretty hard on the innovation front. Everyone (except IE) has fast JS engines now. It's not really a differentiator. They aren't really doing anything notable long term, they're just gliding on their inertia. At least Google is actively working on Mac and Linux ports and extensions for Chrome, which are their two big catchup areas. Mozilla isn't even thinking about process-per-tab.
I would like process per tab, but I like my Firefox extensions more (Scrapbook, Clear fields, No Script, Its All Text, Errorzilla, Download Helper, Add to Search Bar, Web Developer, No Squint, Search Status).
You're right, though, when you say that it isn't just about the javascript engine's speed. So maybe we should wait until Chrome becomes a full-featured browser before we start touting its unbelievable performance. Remember when Firefox went 2.0 and one of the biggest complaints (that I remember) was how it had become bloated and wasn't as fast? Guaranteed the same will eventually happen to Chrome.
There wasn't a whole lot of change from Firefox 1.5->2.0, Firefox 3.0 was the big refresh. I don't actually remember that much complaining about poor performance in Firefox 2.0 relative to 1.x.
What do you feel are missing features in Chrome that actually matter to people that you feel with materially impact performance significantly?
Because, once again, a bunch of software developers have forgotten to obey the %*&(ing theme. Chrome's URL bar successfully takes my foreground colours and completely ignores the background colour setting - so I get a white hostname and a light green path on a white background.
And Firefox 3's awesomebar and the way it handles self-signed SSL certs both make me incredibly happy - maybe chrome does these just as well, as I say I can't tell yet (FVO "can't" meaning "choose not to fill the prerequisites in order to be able to", of course :).
But it strikes me that there are probably a lot of issues like the one I'm complaining about that don't matter to most people but are a showstopper for -somebody- - and firefox has probably nailed almost all of those by now, either in core or in an addon.
Seems to me that the key point of chrome is to provide a decent amount of testing for the tech that's going to go into site-specific browsers for various google applications - really, truly going head to head with firefox doesn't really seem to me to achieve anything for google when -somebody- crushing IE is likely more important to them than specifically them doing it and firefox already has a head start.
I'll tell you why: Google has lot's o' cash and just might start paying OEMs to install Chrome in place of IE. That's something that Mozilla can't do (at least, not on the same scale as Google), and most people use the browser that's installed for them anyway. So what should Mozilla do?
Well, here's a thought: once Google starts paying OEMs to install Chrome, Microsoft is going to have to take a long, hard look at what they're doing in the browser space. Once they do that, they'll find there's no compelling reason to keep draining resources into the black hole that is Trident. That leaves them two choices: WebKit or Gecko. WebKit might be the obvious choice, since it would bring them close to feature parity with Chrome, and isn't controlled by an ex-arch-nemesis.
On the other hand, the Mozilla foundation could write a couple of nice, soothing, "I'm sorry for that whole anti-trust litigation deal" letters to Ballmer. That way they could swap one sugar-daddy (Google) for another (Microsoft), give Microsoft the rendering engine they so desperately need, and continue to develop XUL and Firefox off on their own.
That is totally never going to happen. Gecko and WebKit are both totally useless to Microsoft, because they're good for nothing except browsing public internet sites. The last time Microsoft upgraded Trident's standards support (with IE7), they got bitten badly by all their business customers' intranet sites breaking. With IE8, they've pretty much promised to include the IE7 revision of Trident with every future release, so that none of their paying customers ever have to upgrade their intranet apps ever again.
As for pouring money into the black hole of Trident, they've already stopped that - I believe the new 'standards mode' engine of IE8 is a wholly new contraption, unfettered by all the legacy quirks that bog down WebKit and Gecko.
It's worth remembering that Google doesn't care to win this war - they just want faster browsers to drive web and webapp usage.
There is also greasemonkey support in safari (with greasekit) and opera (no extension needed).
Thinking further, most GM scripts only operate on the html received, and are independent of the browser. So all one needs is the GreaseMonkey platform to act as a holder, and existing GM scripts can run straight off, without modification. (maybe some use special Firefox functions, or their job is to interact with firefox specifics, but I think most wouldn't).
As for Adblock and Greasemonkey, neither plugin is particularly complex. Once plugins for Chrome get big, I'd give it 3 weeks before we see a working version of each.
In your second paragraph, I think you're underestimating the power of a platform (which enables MS to make so much money BTW). It's an incredibly powerful competitive advantage. Remember monkeyboy: "developers, developers, developers".
I agree Adblock and Greasemonkey can be duplicated in negligible time (it took 7 iterations to get the latter's security model right, but that work is now already done) - but that's only 2 addons. How many man-hours to duplicate a comparable eco-system of addons? Secondly, Greasemonkey is itself a platform - how many man-hours to duplicate all the add-ons for that?
It's not enough that there's the ability to create the addons - they also have to be created, and that takes time, during which more addons are created and refined for Firefox... I'm not saying it's impossible to catch up, just that it an established platform is a powerful factor. BTW: Chrome isn't even released for Linux yet. It's fast, yes, but...
An exception is if Google can make Chrome plug-compatible with Firefox plugs-ins; and if a Chrome Greasemonkey can be made that is plug-compatible with Firefox Greasemonkey. From my fiddling around with these technologies, they both seem to be to be horrendously tied to Firefox specifics. They don't need to be, but they are. IMO, anyway. :-) NB: even firefox 2 and 3 aren't plug-compatible! :D
However, perhaps none of this matters, because (I think) few people use many add-ons beyond adblock. But I really don't know. Maybe they do? I guess Firefox's extensions page has stats, so it would be easily estimated. Oh well, why not some data? It's so much more fun to argue with facts to throw:
Adblock: 627,380 downloads weekly. That's well over 25 million pa. mmmm... the currently most popular plug-in is actually "Video DownloadHelper", at 701,065 weekly downloads. Here's the 20 most popular add-ons: https://addons.mozilla.org/en-US/firefox/search?q=&cat=a...
* Major look and feel changes in 3.0
* Spell checking in 3.0
* Performance improvements throughout the browser.
* Improved support for adding/removing plugins
* Allow searching of addons from within Firefox
* Improved Mac OS X integration.
* The Awesomebar.
* Completely revamped everything to do with bookmarks.
* Improved options window
* Revamped Downloads window, with download progress in the bottom right.
Coming soon in 3.5: Private browsing, dragging tabs into new windows. For a lightweight browser, this adds up.
But I cannot accept that Firefox eats up 2g memory after some time and stops responding all together once in a week and kills all my tabs with it. What Mozilla should really steal from Chrome is the one-process-per-tab and sandboxing ...
Has anyone noticed that other browsers using Gecko as their rendering engine (Epiphany, Galeon, etc.) are much faster than Firefox? AFAIK, these browsers are built using native toolkits (GTK+ for both Epiphany and Galeon) of whatever platform they work on. Also, has anyone noticed how slow Songbird is? That's XUL for you.
I don't care about extensions. IMO, a web browser is supposed to let you view and interact with webpages, nothing more, nothing less. Leave the ad blocking, note taking, email checking, photo uploading, iTunes controlling to separate applications. How difficult is it for Mozilla to just wrap up Gecko in a Cocoa shell for the Mac, GTK+/Qt shell for Linux and whatever toolkit is used these days on Windows? Why build the UI using XUL?
I'm an Opera user, but I hate the way they've been piling on feature after feature into the browser. BitTorrent? Speed Dial? If they don't stop now, I'll switch to something leaner.
And why does XUL have to be slow? Couldn't it be sped up instead of throwing it away?
What happens when Google doesn't need that?
But the point remains - It's a lot of money.
I'll stick with firefox if it really does reduce resource usage in its next iteration. But I've heard those kinds of claims before, and nothing has come from them.
- All my FF extensions are available in Chrome & - The ability to change what sites it's saves in my landing page. I don't want to be reminded I got RickRolled.
I switched from Opera to Firefox, despite Firefox being noticeably slower, because of extensions. I love extensions. In fact, right now I'm unable to imagine a web without extensions. It would be like using a OS without the ability to write scripts or code to interact with it. I realize this probably doesn't apply to all Firefox-users, but I seriously suspect it applies to almost all early adopters.
Why would I switch to to Chrome with extensions lacking, when it offers me nothing more than Opera? In fact Opera offers more than Chrome, is just as fast (based on actual machine-to-human metrics, not artificial millisecond wanking) but that didn't stop me from going to another browser.
Speed is only one factor in the equation, and on modern machines it is highly irrelevant. If speed was that important to everyone, why didn't Opera turn out to be the top dog years ago?
* Adblock plus (with Filterset.G) * Firebug * Gamil notifier * Greasemonkey * Platypus (reformat and edit webpages live, script it, ready for Greasemonkey) * Rikaichan (Japanese dictionary with lookup directly in your browser with a simple mouse-hover) * Linkification (Create links out of any non-linked address on pages) * Mouse Gestures (Not just navigation, things like image-zooming and whatever you feel like) * Ubiquity (which pretty much has "extensions" on its own) * User agent switcher (Because some sites gives you different content than google)
Some of these things can surely be implemented in Chrome or will be implemented sooner or later. Some can probably be implemented outside the browser, but that means another layer of indirection while working with your data. In other words: A step back.
I don't expect all of these extensions ever to be ported to the core of any browser. In fact that would make very little sense as only a fringe share of the market would need it or want it.
However having the ability to have extensions and go crazy with your customized install is what makes Firefox great. And this is where Chrome fails. It wont let you make your browser yours.
Firefox may be dead simply due to Mozilla not having any long term strategy vision about how to stay relevant. The issue isn't that Chrome doesn't have extensions now, but what is Mozilla doing now to plan for the future when they no longer have the extension advantage. So far it looks like nothing.
Even Firefox's vaunted new JS engine in 3.5 isn't going to be turned on for extensions, only web content. Meanwhile not only does Chrome's extension code get the V8 love, they also run as a separate process, which means they won't impact performance of browsing as much as Firefox's extension framework does.
I'd call it loyalty for all they've done to the community and the web in general, and I am grateful for that.
Rock on, foxy, rock on.