Firefox 3.1 targeted for year’s end
Mike Shroepfer, Mozilla VP of Technology announced today a draft plan for Firefox and the Mozilla platform beyond the upcoming Firefox 3 (and attached Mozilla platform 1.9) release next June.
Most notable is Firefox 3.1, the next update to the Firefox 3 branch. It would add a few features that were not ready to ship in Firefox 3 development time frame. Among the most relevant:
- Support for the <video> as defined in the HTML 5 specification. Chris Double has been working on this for about a year now and there are even a few experimental Firefox builds featuring it with native Theora (Ogg) support. Whether or not a certain codec must be part of the specification was the subject of a lengthy discussion last year. Sadly, one of the reasons to oppose Theora as a mandatory codec is the fear of a submarine patent that could make browser vendors vulnerable to a patent lawsuit.
- Support for cross-site XMLHttpRequests (XHR) which would allow more powerful web applications and an easier way to implement mashup. Support for cross site XHR was pulled out of Firefox 3 code due to ate changes to the specification.
- More power for Firefox 3’s location bar. Mike Beltzner, Firefox 3 lead, talked recently about how the search and location bars could be merged. In an intuitive way. SeaMonkey -and the Mozilla suite before- already does this but the implementation is not the most discoverable.
- More performance tuning, better system integration.
- Native JSON DOM binding, a powerful feature for web developers.
It would branch from the Mozilla2 code (known as mozilla-central, in the works for eight months already) sometime this summer, in Mercurial, the new version control system Mozilla is moving to from CVS.
Firefox 3.1 would be targeted for this year’s to intentionally coincide with Firefox Mobile (Fennec) development and release, making it the fastest update in Firefox history. It usually takes about a year between releases.
Firefox 4 is targeted for late 2009 (back to year long development cycles) and would introduce Mozilla2, an extensive update to the Mozilla platform to feature highlights like ActionMonkey, the merge of Mozilla’s JavaScript engine (SpiderMonkey) and Tamarin, Adobe’s JavaScript virtual machine open-sourced in late 2006.
Subscribe RSS
Subscribe email
May 20th, 2008 at 2:38 am
Yeah, I already noticed a couple of days ago that the Firefox4 nightly builds changed into Firefox 3.1.
[Reply]
May 20th, 2008 at 5:01 am
Does it mean that Firefox3.1 will pass Acid3?
[Reply]
frostieMay 21st, 2008 at 2:41 am
Who cares?:D
[Reply]
May 20th, 2008 at 3:22 pm
Good news, but better CSS 2/3 and XHTML support would be a nice enhancement too.
Where is the html/css web developer fraction in the Mozilla team?
[Reply]
May 20th, 2008 at 5:55 pm
I hope it passes the Acid3 with flying colours (if you know what I mean).
That should be a priority, but a huge one. Speeding up the loading, of bookmarks and the places when you start Firefox should a high priority. Its quite slow when you open firefox.
[Reply]
May 21st, 2008 at 11:46 am
This is great news. I’ve always wanted the search bar and address bar to be combined. Hopefully it’ll pass on the Acid 3 test
[Reply]
May 21st, 2008 at 11:50 am
I wonder what the filesize now is compared to previous versions, surely it takes up more system resources
[Reply]
May 22nd, 2008 at 4:15 am
Cute.
What was the target date for 3.0 again?
And obviously even 3.0RC is still far from being ready: it just “forgot” ALL of my bookmarks. Not good.
[Reply]
May 22nd, 2008 at 2:23 pm
I think FULL support for CSS3 and XHTML/HTML5 would be the most I could ask for as a web developer. The JSON/DOM and XHR sounds awesome too!
[Reply]
May 23rd, 2008 at 10:39 am
Actually one of the main points of Firefox 3 is improved memory management. Better performance and lower resource requirements have been there since beta 1.
[Reply]
May 23rd, 2008 at 4:18 pm
What I would enjoy seeing is some XHTML 2.0 support since HTML5 is being worked on. Of course, a lot of things are unstable with XHTML 2.0, whereas HTML5 is pretty solid already. I suppose that is the reason why none of it is implemented. Of course, facilities exist to do it. It just isn’t natively implemented the way that MathML or XLink are.
[Reply]
July 3rd, 2008 at 6:17 am
haha already using 3.1 =P just great
[Reply]