Mozilla Launches Firefox 4 for Android
blog.mozilla.com
blog.mozilla.com
The mobile tabs are awesome. And I really like the fact that I can get to the address bar while being half way into the page. The Awesome Bar is also, well, awesome. I thought I might not like it, coming off of an iPhone, but it actually works really well.
My only complaint so far is that spell checking does not appear to be turned on, and capitalization does not work by default. Very frustrating when trying to post comments to HN...
Set it to the default browser in a heartbeat.
http://ftp.mozilla.org/pub/mozilla.org/mobile/releases/4.0/
Windows http://ftp.mozilla.org/pub/mozilla.org/mobile/releases/4.0/w...
Mac http://ftp.mozilla.org/pub/mozilla.org/mobile/releases/4.0/m...
Linux http://ftp.mozilla.org/pub/mozilla.org/mobile/releases/4.0/l...
Really fast though.
- Tabs are actually usable.
- With Sync, you can walk away from your desktop and have your tabs (and history/bookmarks/passwords) on your phone.
(Disclaimer: I work at Mozilla, but have nothing to do with mobile.)
Can I install a sync server on my own domain?
Source code: http://hg.mozilla.org/services/sync-server Instructions: https://wiki.mozilla.org/Labs/Weave/Sync/1.1/Setup
Edit: Also the link seems to be outdated.
http://tobyelliott.wordpress.com/2011/03/25/updating-and-dep...
The PHP "minimal" server is going away in the future, as the Python server becomes better-supported and easier to set up. For now, however, the minimal server is usable and current; the latest version was released just last week.
(The Python implementation will eventually replace the PHP implementation in production as well.)
Try this: http://support.mozilla.com/en-US/kb/what-firefox-sync#w_what...
There is absolutely no way for Mozilla to get your data, even if subpoenaed. And, if for some reason you still don't trust it, you can easily set it up on your own server.
There needs to be more assurances like this in the world.
The password is sent to the server, but the encryption key is not. For more info, see: https://support.mozilla.com/kb/where-can-i-find-my-firefox-s...
What does that mean exactly?
I don't understand why these tabs are better than the tabs in the normal browser. Let me guess, everyone thinks you have to scroll up to get to see the tabs? Too bad.
If so, it will be the first mobile browser to support it.
Basic test here: http://html5demos.com/contenteditable
or much more advanced here: http://ckeditor.com/demo
Others might not appreciate it but that means it's the first mobile browser that rich, WYSIWYG editors like CKEditor and TinyMCE will finally work!
(iPhone and native Android do NOT support contentEditable despite being a webkit branch)
This is history repeating itself where Mozilla/Firefox competition will force other browsers to finally improve their game.
(Remember when IE had 90% market share, and web developers routinely wrote IE-only code?)
Sure, firefox keeps IE in line, but mozilla also has a history of jumping the tracks without strong competition.
The first version of firefox was in 2002, and was already better than IE: http://ftp.mozilla.org/pub/mozilla.org/phoenix/releases/0.1/
Firefox AND Opera made the web bearable in the years between IE6 and Chrome, but they hardly are responsible for the arms race that is taking place now.
The thing I think you're missing in the evolution of IE after the team was disbanded post-IE6 is that it takes time to build up a project team from 0 to hundreds of people, especially when the codebase is preexisting and large. IE7 was them starting to put a team together again, doing catchup on obvious UI features like tabbed browsing. IE8 was them finally having enough people to execute well on some things (CSS 2.1, process-per-tab) and try to play catchup on other things (JavaScript performance). IE9 is them having a team big enough to work on all parts of the browser at once.
You also have your history slightly wrong on the JS perf. Chrome was first announced on September 1, 2008; before that some people knew that Google was working on a browser, but the details were secret. The WebKit project initially landed the SquirrelFish Extreme jit on September 18, 2008; it's pretty clear that this was in the works before Chrome's existence was announced. Mozilla landed the initial Tracemonkey implementation in their main development tree in mid-August 2008 (weeks before Chrome was announced), and had been working on it for some time before that. So all three (Firefox, Safari, Chrome) were working on JITs in parallel before Chrome was announced; it's not like Chrome shipped a JIT and everyone else suddenly decided they had to do it as well. Had Chrome never existed, Firefox and Safari (and Opera, which joined in on the game in early 2009) would still have competed for faster javascript; IE9 would still have ended up doing a jit, or risking being left in the dust.
Oh, and the Sunspider benchmark, for example, was first released in December 2007, when both WebKit and Mozilla were actively working on their interpreters' performance and starting to work on the JITs they'd have by September 2008. Again, the competition for faster JavaScript was very much in play already.
Now Chrome has contributed to the competition, as have all the other players. But they sure didn't start it.
In fact, the Apple WebKit folks have said that they began the Squirrelfish Extreme project the day after Tracemonkey was announced. It was quite an engineering accomplishment to have an initial stab at a JIT in a month and have it shipped to users in a few more, especially considering it took Mozilla about a year (Summer '08 - June '08) to ship theirs. And everyone knew about Tamarin's development over the previous couple years - so it ended up being a really amazing coincidence that all three JITs were announced within a one month period.
- supports <video> and <audio>
- fixed-size divs (used for multiple scrollable page sections) work properly
- Pages are rendered practically identically to desktop Firefox/Chrome; the Android browser sometimes does weird things to pages with complex CSS.
While I haven't noticed any improvements in page rendering speed, but the browser UI itself is more responsive.
EDIT: https://wiki.mozilla.org/Mobile/Platforms/Android#System_Req...
It is the CPU generation. "Release builds of Firefox require a phone with an ARMv7 processor" and the Optimus has an ARMv6.
Is anyone from mozilla or otherwise familiar with how badly it would run if I built it myself with a v6 target?
Building for ARMv7-only lets us use the Thumb-2 instruction set, which results in code that is smaller, faster, and uses less memory.
We used to have ARMv6 builds available, but at this point the browser is not really optimized enough to run well on older (or low-end) devices. And since we were not actively testing or supporting those builds, the code now has some crashing bugs (probably related to the JavaScript JIT) that show up only on ARMv6 builds. Without fixing those bugs, you can't actually run Firefox on ARMv6.
We continue to reduce memory and CPU usage, and if it it gets good enough to support these devices then we might start working on it for a future version.
I'm just having trouble reconciling their incredible benchmark numbers with how it feels to simply browse the web.
http://blog.lassey.us/2011/03/29/mobile-firefox-performance-...
(These performance tests are based on a JS library to do efficient dictionary lookups; more info at https://github.com/mckoss/lookups)
In other areas, like startup speed, we know that Firefox for Android is still catching up to other browsers. It's improved a lot over the last year, and we'll keep on improving it.
You can also search for "Firefox" in the Android Market on your phone (connect to WiFi if possible, if you want avoid cellular data charges).
(By the way, if you literally ended up in a redirect loop, please let me know so we can fix it.)
No "HTTP redirect", I meant navigational/logical/mentally.
For Google Reader, you don't need to change the useragent; you can just go to http://www.google.com/reader/i/
Reader works well in Firefox, but many of Google's mobile sites use WebKit-specific markup, and will not work correctly in Firefox if you change the user-agent header to iPhone or Android.
After that, it just hung on the Loading screen sucking up 100% CPU. I searched to see if I could manually delete the extensions, but they aren't saved on the SD card from what I could find, so I uninstalled and reinstalled the application and now it actually loads.
They really need to put some safeguards in place to make sure extensions don't crash the browser like that. Having to uninstall and reinstall just to remove a bad extension is really problematic.
We're working on various ways to reduce the size, including moving the locale files into separate files that are downloaded at installation time.
Firefox did not support gestures. It was very slow to start. It did not reflow text (properly or at all?).
At least the reflow seems to be included now so I guess I try again.
edit: Installed it. Tried zooming in on reddit, hn and metafilter. Failed to reflow on all of them. Reflowing after zoom is enabled in the options, at least I think it is (the stupid apple slider shows the yes in blue, I guess that means it is on. Or does it switch on if I press it? I never got the design of those buttons.)
Apparently no way to exit the application.
Takes guesstimated 10 seconds to cold-start. 3s for Dolphin.
I'm not sure if Firefox Mobile has the same support for theming that the desktop does, but if so I imagine they'll be some nice themes for it eventually. :)
It's of little use for me without that.
http://blogs.msdn.com/b/ie/archive/2011/03/28/browser-power-...