Mozilla has released the first beta release of Firefox 3.6, which comes with some nice Windows 7 integration features. More specifically, the Firefox 3.6 beta integrates with the new taskbar in Windows 7.
What this means is that tabs within Firefox will show up as previews when you hover over Firefox’ taskbar entry in Windows 7, much in the same way Internet Explorer 8 already does it. This makes Firefox one of the first high-profile applications (that I know of) to offer integration with Windows 7’s new taskbar.
Firefox 3.6 also comes with visual previews when you use the ctrl+tab method to switch tabs – all in a very Windows 7ish coating. This feature is disabled by default, so you need to dive into about:config, find the browser.ctrlTab.previews key and set it to true. Restart Firefox, and experience the new tabbing feature in all its glory.
The new tab switching feature has an additional nicety which probably comes from Andy Razkin: apart from showing visual previews of the tabs, the new tab feature also shows an option called “Show all n tabs”. This will bring up a permanent preview window with an instant search field to search through your open tabs.
Firefox 3.6 also comes with performance improvements, but for the most part, it’s an evolutionary release. On the Mac OS X side, the roadmap for 3.6 includes native dictionary integration, services and AppleScript support, and Keychain integration. The roadmap has nothing to say about Linux-specific features. Get it from Mozilla’s FTP server.
A great deal of work has already gone into Ubuntu gnome integration for Firefox 3.
sigh
Sigh
Ubuntu != Linux
Integrating FF into Gnome on Linux is all very well. Two points come to mind
1) Is this integration being fed back into the relevant products code base?
2) What about KDE, XFCE etc?
..
Gnome != Linux ?
KDE != Linux ?
XFCE != Linux ?
Ubuntu did some integration with Firefox 3.
Squabbling over what defines a Linux is just pointless.
It’s open source. If upstream want it they can just get it.
I guess they’ll have to get cracking on it on their own if they want it.
There is a whole world of difference between making the sources (& thus the changes made) available than actually and pro-actively submitting the changes into the core project.
It is the latter I’m interested in.
If they are doing it then great, fine, fantastic. It is however all to easy to skip this bit.
What? It’s not upstream that has to go around finding stuff to integrate. It’s the ones who create new stuff that have to offer them to upstream projects. This is what it means to be a good citizen in the open source community.
Upstream projects are just the reference code base, they are not people. The people are whoever contributes.
Please, if someone doesn’t share your exact definition of freedom, that does not automatically mean they are not good citizens. It’s a simple as that.
One could also say it’s much more complicated than that. Do you demand from a child that he/she gives back to society the instant they are given something? Does it matter *how and what* that individual gives back eventually? Who will decide what the “giving back” is ultimately worth?
The “open source community” does not have a strict definition for what it means to be a good citizen. You may, however, and i can certainly respect that. Just don’t make it sound like you represent the whole “open source community”.
Edited 2009-10-31 16:49 UTC
If you don’t want to contribute back to the original project, and the license doesn’t force you to (few do), then you don’t have to. You are not even forced to improve the software at all.
However, if you do make and distribute changes but don’t try to merge them upstream, you are not doing anything wrong, but you are also not being a good citizen either. You are taking what upstream gives you, but you are only making the improvements available to your sub-community.
Expecting upstream developers to go around looking for stuff to integrate is just like saying “we take your stuff for free, but if you want our improvements you get off your ass and come get them”. Doesn’t sound nice, does it?
Wrong. You’re making them available to anyone that’s interested.
Oh come on, don’t make it sound like a herculean task. If the changes are good and people want them it won’t be hard to find them
You’ve described one of the reasons why so many businesses avoid open source software like the plague. There’s formally-defined license terms, any reasonable business person would assume “I’ve complied with the terms of the license, so everything’s A-OK.”
But NOOOOOOO, there are these additional unspoken obligations that exist only the minds of the “community.” Despite being vague and undefined, hordes of angry GNU/Freetards will rake anyone over the coals if they don’t meet those imaginary “obligations.” And when you boil it down, most of the “obligations” are along the lines of “give me anything I want for free, and by the way what’s taking you so long?”
So you have a community where the collective sense of self-entitlement is only exceeded by its collective anti-commerce mentality. Most businesses take one look, and run in the opposite direction (as they should).
I like how Android took the linux kernel and replaced everything else with the apache license. It may very well be the future of linux (at least in the mobile space) as it sidesteps GTK+/QT, KDE/GNOME and all the other fighting between thousands of linux distros.
That should be nothing unusual to businesses, though – the differences between ‘legal’ and ‘socially acceptable’ extend far beyond the Open Source communities. Dubious expenses claims by politicians, for example – they might be within the letter of the law, but the public don’t find that to be much of an excuse. Or the tax avoidance that several NZ banks are getting beaten up over – legal, but unacceptable to the public.
Except that Freetards will go into torches-and-pitchforks mode over actions that no reasonable person would consider “unacceptable.” Witness all the whining about Google, or Mark Shuttleworth, etc, because they’re supposedly “not giving enough back to the community.”
The lesson there is: you can spend millions (billions?) of dollars and man-hours on efforts that benefit the “community” – but if you’re a business and you don’t bend over backwards to appease their every single self-indulgent whim, then be prepared to be labeled as evil.
Nobody is morally forced to contribute back, neither companies nor individuals, having or not having the skill to do it. Those are ideas that you and others are trying to read in what I originally wrote. Not contributing back is just fine, as long as the license allows it.
I’m saying that if you do publish your changes but can’t be bothered to push those changes upstream, then upstream developers also have no obligation to go around wasting the time they have to, you know, actually develop, searching for stuff that you could easily bring to their attention.
Implying that all the burden should be on upstream developers, and that it’s their interest to incorporate your changes because you don’t want to spend 10 minutes writing an email and creating a patch _is_ bad practice. They are not your servants.
I don’t follow the cult of Stallman, far from it. If you don’t want to contribute, fine. But if you try to push all the work to those that are creating, for free and in their spare time, the stuff you use, that’s freeloading.
Trying it right now, but the quality of the previews left much to be desired when you compare them with IE8.
Edited 2009-10-31 00:04 UTC
Chrome 3 already supports jump lists although I’m not sure if it supports the new download progress indicator in the status bar icon as well (I’d have to boot into 7 for that and I’m too lazy to do that right now). If you count Chrome as a major application than technically it predates Firefox with its 7 integration. Chrome definitely does not support this tab preview feature like IE8.
Firefox will feature full Windows 7 integration with 3.7. The missing features that didn’t make it for 3.6 (jump lists, progress indicator, DirectWrite support, new Glass theme) are already in the pipeline.
I hope they also bring Direct2D support to Firefox so that all the UI is fully GPU accelerated rather than relying on the deprecated GDI/GDI+.
Cairo recently received an OpenGL backend first since Gecko is supposed to run on platforms other than Windows as well. The Mozilla wiki mentions that the developers are considering a Direct2D backend should the OpenGL performance turn out to be suboptimal.
https://wiki.mozilla.org/Platform/GFX/HardwareAcceleration
Very interesting link. Thanks for posting.
They can use Aero which is accelerated by default if I’m not mistaken.
Linux theme changes for 3.7 and 4.0 https://wiki.mozilla.org/Firefox/Projects/3.7_and_4.0_Theme_and_UI_R…
Also, the Windows 7 integration adds the download progress in the task bar button like IE does. Actually surprised they managed to do this so quickly, OS integration issues usually take ages to go in.
In 3.5.3 you can already see the download progressing in the taskbar icon.
The roadmap must be out of date as neither the keychain nor the dictionary integration are present in the first beta. As far as I know the roadmap was conceived before Mozilla decided to switch to a fast iterative release schedule so these kind of features probably slipped to a later release. I’m surprised though that nobody has done anything about those butt-ugly Windows 95 dropdown lists which have plagued the OS X version of Firefox for years, especially since the managed to make the Windows counterparts look native.
I dont understand what “drop down lists” you are referring to? Firefox on OS X looks good to me, other than that the default theme is horrible IMHO.
This theme should be the default look IMHO;
https://addons.mozilla.org/en-US/firefox/addon/12610
Also, smooth scrolling in firefox is a joke compared to safari. I wonder if that is because of OpenCL?
Smooth scrolling has sucked for me on both Windows and Linux for a long time, I don’t see why would it be different on OSX
A better name for “smooth scrolling” would be “imprecise and unresponsive scrolling.” It’s a crappy stopgap from the days when most browsers/computers were incapable of smoothly scrolling by one line at a time.
The ones you can create with the <select> tag. Compare the lists at the bottom of the following (under display options) under Safari and Firefox and you’ll see the difference:
http://ubuntuforums.org/forumdisplay.php?f=11
And completely lose any kind of visual identity by aping Safari? That’s a no go, Firefox wants to be recognized while still looking like part of the host system. There’s a theme revamp for the OS X version underway though:
https://wiki.mozilla.org/Firefox/Projects/3.7_and_4.0_Theme_and_UI_R…
It used to be under Firefox 3.0 and earlier versions but since 3.5 it works fine on my computer. Can you link to a page where Firefox’ smooth scrolling performance starts to deteriorate?
And no, it’s probably not due to OpenCL since Safari’s scrolling performance has been excellent for years, long time before Snow Leopard got released.
I’m fine with that as long as they keep the ‘small icons’ option to disabled the integrated back/forward button and I can get the bookmarks-toolbar out of the way. I’m used to as small and as little as possible buttons since Netscape communicator and I’d love to keep it that way. 😉
Yeah, that looks pretty bad. But I just tried that page with Opera 9.10 and it looks nearly as bad as Firefox does. Then, just for grins, I tried it with defunct IE 5.2, which I expected would look horrible, but it’s actually better than Firefox and Opera (well, the style is the same, but the drop-down lists have both “up” and “down” arrows like Safari’s does).
Note: I’m running Panther on an aging PPC Powerbook (the most reliable laptop I’ve ever had).
Ah, yes, now i see. They do look kind of odd!
Yes, i happen to think safari did it right (TM). But those mockups look much better than the present firefox, indeed.
My bad, don’t know where i got that from. I guess since smooth scrolling in windows/linux lags for me, i assumed i had tried the firefox “smooth scrolling” native to osx as well.
Actually, it’s much more smooth in OS X. What a pleasant surprise!
Edited 2009-10-31 16:18 UTC
Firefox is not the first “high-profile” application to embrace Windows 7 Superbar. Apple had already done it long back, with iTunes 9 and Safari 4. http://www.beingmanan.com/wp/2009/09/apple-embraces-windows-7-super…
Don’t you mean Aza Raskin instead of Andy Razkin?
They have already added KDE4 file dialogue integration. You can see a video of it here. http://www.youtube.com/watch?v=Pyp5bX_NV6k
I think other distros will implement it soon, hopefully.
Wonder if Firefox integration works correctly with external links. The proper behaviour would be that if the link points to an image file, the picture viewer is opened. If it points to a PDF file, the PDF viewer is opened. And if it points to a web page, only then the web browser is opened. Currently it’s easy to make Firefox KDE’s default browser but that breaks up this viewer-app-by-content feature.
I’ve been saying for a long time that Firefox is basically a Windows-only browser. And this just strengthens my convictions. Sure, there are versions of Firefox for other platforms, but their integration is lousy compared to the Windows counterpart.
The two main things that bother me about Firefox on the Mac, for instance, are CPU usage and interference with power management. Given the same idle state (no animated images, no flash, blocked ads), Safari will use 0-1% CPU, while Firefox will use 5-15% (depending on how many documents are open). Also, there is a bug in Firefox that’s been known for YEARS that prevents a Mac from going to sleep after a period of inactivity. Both of these are hostile to notebook users especially, wasting energy and shortening battery time.
It just amazes me how QUICK they are to add features for Windows but how totally disinterested they are in every other platform. On the one hand, they want to come off as though they’re all about Free Software and multi-platform. But when it comes down to looking at their actions, it’s clear that all they care about is displacing IE on Windows. That’s is. Everything else is merely tolerated.
Sadly, I tend to agree. It took them forever to change the form buttons to match the GTK theme that you’re running under Linux.
But in their defense, Windows still does take up 90% of all workstations.
At this point, I just think it’d be sweet if Firefox would be ported to my Amiga 4000. But unfortunately I think it’d eat too much memory!
They’re putting effort into the majority, but the priority lists seem to indicate that they’ve put less emphasis on Windows with 3.5.x.
After using Adobe software, an idle CPU usage of 1.x % would be nice. For so long, software with any legacy has been using the old loop and poll for a message structure. As far as I know, only the NeXT vendors like OmniGroup are doing it really well, as third parties go.
Many people seem to complain on Mac OS X about a native look, but what they really mean is that it has to look like Safari and that’s just not right. If Apple had never created a browser, would Firefox and Opera have to look like OmniWeb?
My only serious complaint is about the pages and the combo/list boxes. The people behind Firefox have admitted to doing a bad imitation and while Opera has a better imitation, Chromium seems to use the native widget.
AFAIK even Safari Webkit only emulates native Cocoa widgets in a web page for performance reasons. It’s just that Google seems to put a lot more effort into getting Mac Chrome right from the start.
Edited 2009-10-31 17:23 UTC
I’m sure that has something to do with Chrome being the platform for Google’s office applications and their desire to make things as simple and seamless as possible for users.
I just wish some of that thinking would make its way back to Firefox.
Well, this is as good a time as any to note that Epiphany-Webkit in Gnome 2.28.x is finally ready to rock. Complete with extensions… including adblock. It’s lean. It’s mean. It’s fast. It’s powerful. And it’s totally dissociated from Mozilla Corp and ponderous resource hog, Gecko.
That said, if Gnome is not your cup of tea, there is Chromium, which works quite well. And of course, Midori is also Webkit-based, and a very nice match for lightweight DE’s or no DE at all.
Gone are the days we have to beg for scraps from Mozilla Corp. Asa, I spit upon thee!
Hallelujah!
Unfortunately I have to agree with you. I think the problem goes even further somewhere shortly after the mozilla -> firefox transition the mozilla guys just forgot who where the users who initially adopted them and advocated them. More and more “advanced” features were taken away and development concentrated more and more on Windows only. My prime example for not thinking about their advanced users is the reply-to-list bug in thunderbird (https://bugzilla.mozilla.org/show_bug.cgi?id=45715). It took 9 (!) years to implement this behaviour which I think is essential for people using thunderbird to regularly read mailing lists. I got so fed up with this bug that I stopped using thunderbird. The discussion in the comments is quite enlightening.
Similarly now with firefox I find that it has been regressing on linux for quite some time. I see regular freezes and slowness. And I don’t see a willingness by the developers to address these problems, because it works on windows. I haven’t found a browser which I like enough to replace firefox yet (mainly due to some extensions), but I imagine that I’ll be moving away from firefox pretty soon.
I posted earlier about Firefox vs the new Webkit based Epiphany. But I neglected to provide any actual metrics.
So here we are. Epiphany 2.28 gets a 100 in ACID3 and looks right. Firefox 3.5.3 gets a 93 and doesn’t.
And Epiphany 2.28 totally blows the doors off of Firefox 3.5.3 for javascript. Beats it by an overall factor of 5+. And blows it away in every category.
Opening the browser and going to google.com results in a 48 MB res – shared for Firefox, and 17 MB for the new Epiphany. All the claims of improvement, and Firefox is still a slovenly memory hog.
What is Mozilla Corp *doing* with that $70 million a year, or whatever, that they get from Google? Not building a better browser for Linux, obviously. I’d suggest doing without whatever extension is holding you back and throwing your weight behind a project that cares, for the greater good of Linux browsers.
TEST COMPARISON FROM TO DETAILS
====================================================================== =======
** TOTAL **: 5.14x as fast 1908.6ms +/- 2.1% 371.4ms +/- 0.8% significant
====================================================================== =======
3d: 4.36x as fast 221.6ms +/- 0.5% 50.8ms +/- 1.1% significant
cube: 6.11x as fast 74.6ms +/- 0.9% 12.2ms +/- 4.6% significant
morph: 3.13x as fast 77.0ms +/- 0.0% 24.6ms +/- 2.8% significant
raytrace: 5.00x as fast 70.0ms +/- 1.3% 14.0ms +/- 0.0% significant
access: 9.27x as fast 294.8ms +/- 4.6% 31.8ms +/- 1.7% significant
binary-trees: 7.95x as fast 31.8ms +/- 1.7% 4.0ms +/- 0.0% significant
fannkuch: 9.24x as fast 116.4ms +/- 4.1% 12.6ms +/- 5.4% significant
nbody: 12.5x as fast 112.2ms +/- 4.5% 9.0ms +/- 0.0% significant
nsieve: 5.55x as fast 34.4ms +/- 18.5% 6.2ms +/- 9.0% significant
bitops: 13.0x as fast 233.6ms +/- 4.8% 18.0ms +/- 4.9% significant
3bit-bits-in-byte: 12.5x as fast 30.0ms +/- 0.0% 2.4ms +/- 28.4% significant
bits-in-byte: 9.17x as fast 55.0ms +/- 0.0% 6.0ms +/- 0.0% significant
bitwise-and: 23.9x as fast 91.0ms +/- 12.0% 3.8ms +/- 14.6% significant
nsieve-bits: 9.93x as fast 57.6ms +/- 1.2% 5.8ms +/- 9.6% significant
controlflow: 9.29x as fast 26.0ms +/- 0.0% 2.8ms +/- 19.9% significant
recursive: 9.29x as fast 26.0ms +/- 0.0% 2.8ms +/- 19.9% significant
crypto: 6.15x as fast 108.2ms +/- 0.5% 17.6ms +/- 6.3% significant
aes: 4.45x as fast 43.6ms +/- 1.6% 9.8ms +/- 5.7% significant
md5: 7.38x as fast 31.0ms +/- 0.0% 4.2ms +/- 13.2% significant
sha1: 9.33x as fast 33.6ms +/- 2.0% 3.6ms +/- 18.9% significant
date: 2.26x as fast 153.4ms +/- 0.4% 67.8ms +/- 2.4% significant
format-tofte: 2.49x as fast 68.2ms +/- 1.5% 27.4ms +/- 4.1% significant
format-xparb: 2.11x as fast 85.2ms +/- 1.2% 40.4ms +/- 1.7% significant
math: 6.30x as fast 209.2ms +/- 1.4% 33.2ms +/- 4.1% significant
cordic: 10.2x as fast 77.4ms +/- 0.9% 7.6ms +/- 9.0% significant
partial-sums: 4.98x as fast 95.6ms +/- 3.5% 19.2ms +/- 2.9% significant
spectral-norm: 5.66x as fast 36.2ms +/- 1.5% 6.4ms +/- 10.6% significant
regexp: 13.4x as fast 206.0ms +/- 10.6% 15.4ms +/- 7.2% significant
dna: 13.4x as fast 206.0ms +/- 10.6% 15.4ms +/- 7.2% significant
string: 3.40x as fast 455.8ms +/- 1.4% 134.0ms +/- 1.3% significant
base64: 3.00x as fast 37.8ms +/- 2.8% 12.6ms +/- 5.4% significant
fasta: 3.83x as fast 97.4ms +/- 1.5% 25.4ms +/- 2.7% significant
tagcloud: 3.67x as fast 105.6ms +/- 2.4% 28.8ms +/- 1.9% significant
unpack-code: 3.74x as fast 154.0ms +/- 2.4% 41.2ms +/- 2.5% significant
validate-input: 2.35x as fast 61.0ms +/- 4.3% 26.0ms +/- 3.4% significant
Edited 2009-11-01 00:22 UTC
This is the link which describes the path Firefox wants to take visually for all platforms up to version 4.0 (take it as community service ;-):
https://wiki.mozilla.org/Firefox/Projects/3.7_and_4.0_Theme_and_UI_R…
This site seems to be updated more regularly as well.
what the hell is that. their roadmap transforms the firefox gui into the chrome gui.
are they trying to trick themselves (or us?) into thinking they actually came up with this?
it isn’t a coincidence that firefox 3.7 smoothly mutates the gui halfway to chrome, and 4.0 goes the other half. these changes are distributed well enough to ease current users into the new interface without a culture shock.
it is hackish to not cite chrome as their source of inspiration.
And the problem is? If Chrome UI is superior, they better copy it.
well if all they can do is copy and not innovate in the UI space, then what’s the point? gecko isn’t the fastest and most compatible engine anymore, the only thing it has are it’s plugins which it seems Chrome is going to be solving shortly. They don’t even create the plugins too, so I assume once the framework is there devs will move their stuff over.
of course it’s entirely up to them to not innovate, but you can almost see history repeating itself if that becomes the case and Mozilla corp losing market again.
It should also be noted that Mozilla has lost so many devs over the years and I would guess from their stories a lot of it has to do with shitty management and their Windows focus. Look how Mozilla/AOL people cancelled Camino two days before it was going to be officially launched at Macworld just for the sake of not releasing it and ‘competing’ with Firefox.
They just don’t get the other markets.
firefox was a nice small powerless FOSS project just like every other until google realized it was so superior to the alternatives that they would pay firefox hundreds of millions of dollars to ensure there was a good way to use google and the internet.
google has its own browser now and all the people at mozilla who had their big managerial pockets lined are looking to claw their way up the cliff of despair from the pits of death. as soon as google chrome came out they knew their meal tickets were coming to a stop.
not only is firefox not innovating, they’re pandering. they think that by doing as much as possible the way the master likes, they’ll get some more drops of milk to lap.
the problem it is f–king impolite
Title says it all. I just read that the last remaining Windows 7 feature was pulled after the beta was released due to serious problems with the current implementation which can’t be fixed in a reasonable amount of time (and without further beta exposition).