A lot of work is under way in designing the user interface for the next two releases of Firefox – 3.7 and 4.0 – and both of the currently proposed themes (Windows-specific) look interesting. These interface refreshes were needed as well, as the current Firefox interface is showing its age. Looking at the mockup for Firefox 4.0, it all becomes clear: This is Firefox – Chromified.
Firefox 3.7
Firefox 3.5 is barely fresh out of the door, but we’re already looking ahead to the next release ofeveryone’s favourite browser (except for me, that is). It’s called Firefox 3.7 for now, and the discussion on the proposed user interface overhaul is currently in full swing.
The Firefox team will be embracing Aero Glass on Vista and Windows 7, which is definitely a step forward. Sadly, though, they seem to be going a bit overboard with all the widgets, buttons, and other frivolities that confuse me. The individual pieces of the puzzle look good, it’s just that there are too many different pieces that don’t look very coherent, and are thrown all over the place. Barely any two buttons are the same, which doesn’t strike me as good UI design; there’s also a lot of wasted space. A good thing is that it also works well when Aero Glass is turned off.
Windows XP users surely aren’t left out in the rain by the Mozilla, as the implemented more or less the same design as above to the default Windows XP themes. The theme, as well as the button layout, do not copy over very well to XP’s Luna theme – and I wouldn’t dare to think how it would look on Windows Classic, which is – I surely hope so – the preferred theme on most Windows XP boxes.
Firefox 4.0
The mockups for Firefox 4.0 is where it gets really interesting. As it turns out, the Firefox UI team is experimenting with a Chrome-like tabs-on-top look, but they remain unsure whether or not tabs-on-top is actually the way to go. As a result, they designed two variants: tabs-on-top, and tabs-current.
Both variants again suffer a bit from what I call visual overload – too many different types of buttons, too many different colours, and they’re all over the place. It reminds me a lot of the current Internet Explorer interface, and that’s a bad thing. They are, however, taking a lot of cues from Google Chrome’s interface, but while they can copy certain aspects of Chrome’s UI, they failed to grasp the overall idea: reduce the chrome as much as possible, allowing you to focus on what really matters: content (web pages). The mockups are too busy for my taste, drawing my attention towards the user interface elements instead of towards the content.
Final note
Still, it’s good to see that the Firefox team is looking into improving the user interface of their highly successful browser. Even if you don’t like where these mockups are going, Firefox’ flexibility will most likely allow you to go back to the current interface quite easily.
An issue I haven’t seen addressed yet is how these themes will copy over to the Linux and Mac OS X worlds. I’m sure mockups for these platforms will pop up soon enough.
i like them…anyway they are just mockups a lot can change on the final version
I agree. It doesn’t seem too busy or inconsistent to me. Except for the tabs-on-top change in 4.0, it looks almost exactly the same as the 3.7 mock-up.
The mockups for the 4.0 interface have a bookmark button in the place which is usually occupied by the search box. A hint that Mozilla is considering to implement Omnibox-like functionality for Firefox?
firefox 3.0 address bar already works like an omnibox. Just not by default ;-).
By default its im feeling lucky but its easy to change and easy to use different search engined using their keywords. I for one don’t use search box.
A two-year-old article from Wired, which I agree with: http://www.wired.com/software/coolapps/news/2007/05/firefox_bloat
Personally I try to keep Firefox look & feel maximum clean. And I want my Firefox to be as fast as possible.
By the way, it’s my first post to OSAlert (though I’ve been around for 2.5 years), so hello people
A lot has changed since then – Firefox 3 was faster than firefox 2. Firefox 3.5 is faster than Firefox 3.
As for Google Chrome (that someone else mentioned) – I may consider using it after they add the ability to automatically delete cookies etc upon closing the browser. Until then, I dislike the feeling that my privacy is not important that the lack of such a setting gives me.
While I agree Firefox is really good on Windows, it is still slow as hell in Linux.
Try it both on Windows and Linux:
-Open 10 tabs at once.
-Change tabs while having 20 opened.
-Press Ctrl and make a few full turns on your mousewheel.
-Include Flash, Java, whatever plugins in the previous experiments for 100x effect increase.
Linux is unimportant for the Mozilla foundation, that’s why I stick with Opera.
I ABSOLUTELY agree with you. Mozilla doesn’t care a ***** about Linux, and even less about KDE. They won’t do anything about artwork neither performance.
From my personal experience with Mozilla foundations, they’re far too bureaucratic and self-centric.
While I agree Firefox is really good on Windows, it is still slow as hell in Linux.
I’ve been saying that for ages and ages: I use Windows and Linux on the same machine, FireFox on both OSes is the same version, but still the Linux one runs noticeably slower. It’s not even a graphics card driver issue since I’ve had to change graphics cards several times, both nVidia and ATI, and still the same behaviour occurs. Why? I have no idea, do they even try to optimize the Linux version? Or is it some deeper issue, maybe not even related to FireFox at all itself? Every time I ask some developer about it they either change the subject or just completely avoid talking to me.
I absolutely agree with you.
As soon as distro dvd is out, first thing on my list is removing firefox in favor of any GtkWebKit browser.
I don’t like the feeling of being 3rd rate citizen and the more time passes, the more obvious this is from firefox side. I say… let them have it, there is now enough native browsers in linux which work decent.
Maybe they can count me as download since it came with my distribution, but as user they sure won’t.
I remember reading somewhere that windows firefox under wine is faster then linux native firefox.
I remember reading somewhere that windows firefox under wine is faster then linux native firefox.
I’ve actually read and heard that myself too. Haven’t tested it myself, but I wouldn’t actually be surprised. The speed difference really is painstakingly clear and visible to anyone who uses both OSes. But do we got any FF devs here lurking around? Would be lovely to get atleast some explanation as to why it is like that? X? Some external library? Kernel issue? Or is it FireFox itself?
Old news. Very old news.
This is like the reviews of IE8 beta versus Firefox 2 … written when when Firefox 3.5 was also in beta. Go figure.
Why is it that people complain about versions of Firefox that are over two years old? It has moved on, people. Firefox 3.5 is here, and it is fast (perhaps not quite as fast as Chrome, but almost), and with its use of XUL it retains its abilities for customisation via extensions (without loss of speed) that Chrome has no hope of matching.
Keep telling yourself that.
You don’t work with any large tables do you? For our internal business apps, which do, legitimately, use large tables for reporting, Firefox is a nonstarter. I have to direct users to other browsers. Anything based on Webkit is fine. Opera works very well. Pretty much anything, even IE6, runs circles around FF3.5’s table rendering speed. And as for memory consumption with large tables… FF 3.5 is obscene. 100k of text data should not take hundreds of megabytes to display. But it does with FF 3.5. Some reports push it up past 1GB.
I don’t mean to pick on FF 3.5 here. All the previous versions of FF which I have tested were at least as bad.
Edited 2009-07-28 03:49 UTC
Regardless of whether it works better with your Intranet page, Firefox 3.5 IS much faster in general. If you have a PUBLIC page that is demonstrably slower, then it is useful as a test case or to file a bug report.
Any larger table will cause the problem. There is nothing specific to our apps. And the rendering time is *exponential* as the rows increase. If you prefer to bury your head in the sand and let IE7 under wine walk all over FF3.5 for performance, that’s fine with me. We have other browsers to use which perform just fine.
Edited 2009-07-28 12:27 UTC
I repeat, the performance improvements in Firefox 3.5 are well known. The only one burying the head in the sand is you. You claim a problem exists but refuse to show any link that demonstrate the problem. How about giving one before attacking others asking for more details?
Create a small script that writes a very large table, just grab the data from /dev/random if you don’t have anything better. Test the output in various browsers. Profit?
I want a real world test case. Not a contrived one. A public website demonstrating the problem would be useful. Barring that, a script that generates such a troublesome webpage would do. I am not claiming that the problem does not exist but I am merely interested in getting it reported if it there is a demonstrable test case. Been using Firefox 3.5(.1) for quite sometime in Fedora 11 and as a very heavy user have generally have found the performance to be much better. No doubt about that.
I want a real world test case.
It IS a real-world test case if it affects an entire company. Just because you aren’t allowed to see their private data does not make it less of a real-world one. They could just fill it with random data, but how would that be different from the approach I gave?
I haven’t tried it myself nor do I really even have any interest in trying. I use FireFox myself, and I have noticed it is slower than f.ex. Opera. Opera is like a greased lightning compared to FF. But I like the customizability of FF and there’s several addons I don’t want to be without, as such I’ll stick with it for the foreseeable future myself.
I can claim it affects the entire world but without a demonstrable public test case, is it useful? One needs to be generated first. So a script is just fine if one is available.
You heard it here first, folks. Rahul Sundaram of Red Hat, Inc., enterprise Linux vendor and owner of JBoss, does not consider intranet applications to be “real world” cases, and thus considers them to be unimportant.
Well, Rahul, I’m not going to give you access to our inventory management or ad hoc reporting apps, especially since you don’t consider such things important enough for me to expect you to be careful what you do in them. However, I will give you a python script which generates a test case which exhibits the problem FF has exhibited for many years. Since 1.x, in fact. Though I must observe that there is a special place in Hell for bug triage staff who, when given a simple recipe to reproduce a problem, still insist that the end user do *all* the work.
So if you don’t mind pulling your head out of the sand, and/or out of your ass for a moment, run the Python script below. I’m sure the formatting will be ruined. You’ll probably have to reindent the for loop body.
You will find that while FF 3.5 takes over a minute to render the resulting page on reasonable hardware, Opera, and even IE6 under wine(!), render it in 3 seconds using an order of magnitude less memory.
Still don’t think it’s “real world”? Then you’ve never done any real business reporting apps. Which is understandable, since most web apps on the greater web thus far are toy apps from the standpoint of actually requiring browser capability. Maybe FF 3.5 is snappier for shopping cart checkout and pr0n. But not for heavy lifting in real business apps.
=====
print ‘<html><head><title>Firefox is dog slow!</title></head><body><table align=”center” border=”1″ width=”100%”>’
y=0
for x in xrange(10000):
print ‘<tr>’
print ‘<td align=”right”>%s</td>’ % y
y+=1
print ‘<td align=”right”>%s</td>’ % y
y+=1
print ‘<td align=”right”>%s</td>’ % y
y+=1
print ‘<td align=”right”>%s</td>’ % y
y+=1
print ‘<td align=”right”>%s</td>’ % y
y+=1
print ‘<td align=”right”>%s</td>’ % y
y+=1
print ‘<td align=”right”>%s</td>’ % y
y+=1
print ‘<td align=”right”>%s</td>’ % y
y+=1
print ‘<td align=”right”>%s</td>’ % y
y+=1
print ‘<td align=”right”>%s</td>’ % y
y+=1
print ‘<td align=”right”>%s</td>’ % y
y+=1
print ‘<td align=”right”>%s</td>’ % y
y+=1
print ‘<td align=”right”>%s</td>’ % y
y+=1
print ‘<td align=”right”>%s</td>’ % y
y+=1
print ‘<td align=”right”>%s</td>’ % y
y+=1
print ‘</tr>’
print ‘</table></html>’
Edited 2009-07-28 14:56 UTC
Now you are being plain silly. My opinions on osnews have nothing whatsoever to do with the organization I work for. I am sure you wouldn’t find it fair if someone associated your random comments to the organization you work for like that either. So don’t stop being civil. If you want to be useful and have a test case, file a bug report. Ranting and being rude here changes nothing.
Besides, I never said Intranet websites are not a test case but since I can’t access yours, using your Intranet website as a example is completely useless. I could claim my Intranet website brings down all the browsers except Firefox 3.5 as well. See the point?
Edited 2009-07-29 01:34 UTC
It is much faster in my experience. It is also, however, less stable, especially on OSX. Sometimes I’ll be focusing on another task and I want to go back and look at a web page I was looking at five minutes ago. And Firefox … is nowhere to be seen, save the Ooops Fire fox crashed notification. Oh, well its fine for casual surfing of everyday sites, but for research I switch to Safari.
This is due to firefox’s session restore feature. It is trying to store all the data in the forms elements, so that if the browser crashes, it can restore all the form data along with the page. (using an XPath query for each element which is very resource expensive)
Unfortunately this causes it to shoot itself in the foot when it comes to large tables/forms.
It’s a known issue and is being fixed. (supposedly in xulrunner 1.9.1.3)
https://bugzilla.mozilla.org/show_bug.cgi?id=477564
https://bugzilla.mozilla.org/show_bug.cgi?id=478107
For security reasons, the data is not stored when using an https:// connection, so the problem disappears when coming from an https:// URL.
Session restore might have made the problem worse. But this problem dates all the way back to FF 1.x. Back then we did not have much in the way of alternative browser choices and we just had to suffer with it.
Edited 2009-07-28 12:35 UTC
This is like the reviews of IE8 beta versus Firefox 2 … written when when Firefox 3.5 was also in beta. Go figure.
no one complained for version 1.
but lol, it would almost be better if 2 never came, and since then, linux version has only gone from worse to disaster,
those mockups look almost EXACTLY like opera 10.
swap the tab bar with the navigation bar and it’s almost exactly the same.
I wanted to write the same. Nice to see how Firefox still keeps copying
Apart from that, I don’t really like Opera 10’s look either… I’ll switch back to “Windows Native” as usual.
I have switched to Chrome for about 5 months now. Firefox is not on my system anymore. Only sleeps on my Linux box, yes, sleeps because I am not using it any longer (and it is sadly integrated with Gnome-Ubuntu). It would take much more than UI changes now, for me to deconvert from Chrome. Chrome is faster, and webkit has certainly an advantage on Gecko (faster on my end).
I am wondering here… where the Mo community were going to head if it wasn’t for the brilliant glorious Chrome: Would we be watching the mediocre fight between Mozilla, IE and Opera, stagnant, unable to recreate the browsing idea until these days? Conservative UI’s that fill the entire space of my screen, thousands of options at which I wouldn’t even know what means or where to start from? about:config screen nightmare? Certainly without Chrome, the Mo community would not have moved a finger…until who would know WHEN!
Actually, something like this has to happen to both GNOME and KDE projects.
They seem to be making bookmarks much less visible – no bookmark toolbar in any of the mockups, just a button that I assume drops down the equivalent of the current bookmarks menu.
Not sure I like that – while the fancy address bar is useful, I still use bookmarks a lot in FF3.5 – particularly in the form of javascript bookmarklets on the toolbar, and groups of related links. Nothing I’ve seen here addresses those requirements…
I really hope there will be plugins to enable the classic 3.5 look. I for one do not like Chrome’s UI and do not want Firefox to move in that direction.
I don’t see anything wrong with Firefox’s UI the way it is now (apart from not having a builtin option to disable the menu bar, but a plugin takes care of that).
Everyone always calls Chrome’s UI innovative. Frankly, I don’t see it. What’s so innovative about it? That the tabs are on top? The word innovation used to mean something, now it seems like any little gimmick is hailed as a ground breaking and innovative.
I don’t want a omnibox! No! ANd I want a bookmark toolbar. And a status bar (containing access methods to plugins like stylish, greasemonkey and firebug).
If I have understood it correctly, the idea behind the tabs-on-top concept is that each tab together with browser controls would be better associated with each web page. This is very understandable from Google’s point of view with their vision for the web and their web applications.
However, I think that it would a big drawback for usability since the tabs are arguably the most used feature in the browser interface and therefore should be the most accessible, i.e. the closest items relative to the content and your mouse cursor.
Why put the more infrequently used chrome between the most used chrome (tabs) and the most important (content)?
P.S. I can’t believe they are FINALLY going to combine the Stop and Reload buttons in 4.0! This is the way it should have been done from the beginning… The “hidden menu bar” is also a must-have!
Edited 2009-07-28 00:35 UTC
Well, there’s the Fitts’s Law argument that the edges of the screen are some of the easiest places to reach, since you can’t overshoot them.
Combined Stop/Reload has some drawbacks:
– You have to double-click to reload a loading page.
– If the button changes the instant before you click it, you can end up reloading a page instead of stopping it as intended. This was a bigger annoyance in the dialup days, but still somewhat holds true. (Not to mention some people are still on dialup.
In any case, I’m glad they aren’t planning to put Stop and Reload on opposite sides of the screen like Chrome. Not even IE does that.
Regarding Chrome, the devs (or rather, Ben specifically) flatly refuse to change the default button layout, so vote/comment on Issue 1656 instead, if you care:
http://code.google.com/p/chromium/issues/detail?id=1656
(Provide ability to create custom toolbar buttons)
I’m happy as along as the menu bar, window borders, tabs, and other widgets actually follow the guidelines of the host desktop environment i’m running firefox in.
I like the general idea, but that huge Page button seems out of place. It wastes valuable Tab space.
I agree. The current little pull down box is nice. Not that I use it much, but at least it’s out of the way for the most part.
A bookmarks tab instead would probably be better. It could bring up the bookmark side panel exposing more people to the search feature.
If its customizable to my liking like the old versions im all fine whatever changes they pick.
I like my browser to look like this: http://judgen.googlepages.com/browser.jpg
That screenshot reminds me of opera 3
http://files.myopera.com/tarquinwj/albums/45511/Opera3.png
Tabs on top should not mean removing the title-bar, because now instead of having the entire window width to see the title, you only have a 200px wide tab.
Also you could have no title bar at all, or a 2 pixels high one (on Linux at least).
Title bar and window contents should stay separated, and these tabs belong to window content. If I had tabs on the title bar that would be because I’m grouping windows myself through the window manager.
I really like the green loading bar in the 4.0 mockups.
Good to see that they’re at least shaking things up and soliciting feedback. Part of the reason I haven’t been using FF in the last couple of years is that its interface is plain FUGLY. Happily, Opera and Chrome have forced the Mozilla peeps to give FF some UI love (which Thunderbird also badly needs).
That’s a matter of taste, really. Personally I hate how the custom interfaces of Opera and Chrome don’t match the rest of my OS.
I hope that the final versions aren’t too far from what’s depicted in these mockups, the versions with conventional tabs looks a lot like my current Firefox setup in Win 7, which I can’t be arsed to take a screenshot of since I’m using my Linux box.
If anyone wants to make Firefox 3.x look a bit more like this then they can follow my reminder guide at http://portunus.net.nz/rtfm/bls
Way to go Firefox!
Seriously? A black page with bright green text and dark blue links? Why would anyone take your advice after seeing your bad taste?
Thank you for your profoundly shallow and worthless analysis.
I’d say it’s visual underload. I like the tabs-on-bottom mockup, but I’d like to see a status bar and a full size bookmark bar (it can be autohidden, or off-switchable, I don’t care, but it should be available). I never liked any recent trials to make browser more clean, since more clean ment less usable. Gee, I still fail to understand how these mockups can mean a visual overload to anyone.
This looks really great to me, both Firefox 3.7 and Firefox 4.0 mockup designs are amazing. I personally like tabs on top though. We are going to have exciting times ahead.
One of the things I like most about FF is the customization. That’s why I wasn’t distracted by Chrome at all. I personally don’t like my tabs above or below the address bar; I like them on the left side of my screen. It works really well with wide screen monitors. I also like my address bar up there with my file menu, so I don’t have any wasted space.
I’ll take whatever speed/memory enhancements FF 4.0 has to offer, but I don’t see any of this being much of an improvement to the UI.
I think they all look good except the tabs-on-top, which IMHO is a bit confusing and looks a bit messy. My favourite is the tabs-on-bottom for Firefox 4.0 – really hope they go with that or at least make it configurable.
The tabs on top is actually the better concept, since the address bar is very specific to the current document (your current ‘view’) rather than some meta thingie (like a toolbar, which sometimes does relate information, but often nothing more than is already visible (think ‘bold’ button in some office suite, which is on when you’re in a bold text and off outside it).
But these discussions leave me wondering whether or not we’re still to focused on the “current URI” design. Why not something closer to what some filebrowsers do: A nice representation (which mortal users understand, not just power users). This could be a bit breadcrumb like or pherhaps the website host (which is pretty important information) with the title of the current page. Ofcourse you can simply access the raw URI by clicking on this “titlebar”.
Beter highlighting of the hostname of the website is a good thing as it helps people better understand where they are in general. See all the work in chrome and other browsers since where the hostname is highlighted.
All other information is pretty much too site specific and non-meaningful (often numbers for items and such or simply the title of article (see your uri bar now ;-))
Currently 99% of all my browsing needs would be satisfied by 1 big field with this title/address-bar thing, a back button, stop/reload button, a (optional, I don’t use it on all my computers) bookmark bar and a status/activity bar at the bottom (where i’ll hardly notice it, but where it can display some additional useful information ranging from downloads/ page propertes/ page search/plugin thingies (firebug/ comments/ whatever))
Search may be correctly incorporated in some awesomebar logic, as long as i can simply do 1 click/shortcut websearches.
Forward buttons don’t really add anything for me I think I can count all times I *ever* used them on my two hands but it’s not too intrusive.
What I do very much like about firefox is that it’s context menus almost always contain what I want, which is sorely lacking (too simplistic) or too bloated in a few others.
All in all I think savings can be made (and it should always be very customisable) and that some changes might even make things more useable for the non-tech savy.
But that’s of course just my ^a'not0.02
Firefox 3.5 just plain sucks. It’s the only time I’ve had to down grade a software package. It’s stability and performance are worse than 3.0.
the option to put the navigation buttons on the side panel. This seems, to me, like a natural evolution as more widescreen displays are used. The address bar and tabs could still be at the top, while the back/forward/home buttons are vertical on the side, opening up more vertical real estate for webpages, especially on the smaller screens such as netbooks. my $.02