Wednesday, June 23rd, 2010

Firefox joins Chrome in plugin crash protection

Category: Plugins

Yesterday Mozilla released a new Firefox 3.6 point release including regular stability fixes and a plugin crash protection system that was originally planned for Firefox 3.7.

The crash protection isolates plugins in their own process, meaning that a plugin will not bring down the entire browser when it crashes or freezes. Mozilla states that one of three browser crashes are caused by third-party plugins.

Currently only Linux and Windows include the crash protection, Mac users will have to wait for Firefox 4 planned to be released later this year. This is a fantastic addition in a point release!

Posted by José Jeria at 11:47 am

2.3 rating from 3 votes


Comments feed TrackBack URI

When is Webkit 2 going to launch. it too has isolated crash for plugins.

Comment by MySchizoBuddy — June 23, 2010

1 in 3 firefox browser crashes are caused by third party plugins?
So 66% of all crashes are caused by unfixed firefox bugs?
That doesn’t sound terribly good. Maybe they should be tackling that larger percentage first.
MySchizoBuddy: The two main webkit browsers already have plugin isolation :]
But yeah, will be cool to have it in the reusable library

Comment by ProPuke — June 23, 2010

@ProPuke: Maybe a single feature that covers one third of all crashes and is already implemented and tested in another two open source browsers was easier to add in a point release, than fixing and regression testing hundreds of single misterious bugs.

Indeed, as you surely know, many times debugging and new features have different development paths and teams involved; also, new feats help attract and expand the userbase, where bug fixin’ is less sexy.

Let’s face the truth: I would find a release less sexy (in a sort of unconsciously thrust) if it would claim itself as maintainance only, than knowing that it brings new bells and whistles and zero bug fixes.

However, I can understand that rationally speaking, a software running uninterrupted for more than half a day like Firefox would greatly benefit from a major boost in stability, more than adding features like DirectWrite graphic acceleration or Personas.

But everyone here knows that the problem for the foxes is that the WebKit team is producing great results in really short times: great, stable browsers, cutting edge feats, plugin and tab isolation, and so on. They need to put on some lipstick, a great dress and a good perfume again, or in the long term, the distance could become unbridgeable. Heck, even the IE team acknowledged that.

Comment by giank — June 23, 2010

@giank: You are, ofcourse, completely correct.
I was trying to be half humorous/pedantic.
Although the figure does scare me. Surely the percentage of internally triggered crashes should be considerably lower for something that is no longer beta (or indeed zero).

You make all-good points.
I fear the mozilla+xulrunner framework might have too much weight to effectively react in todays market (or at least keep up with webkit & chromium).
Odd. Agility was originally firefox’s trump card (with the advent of such an extendible application).
Odder still that Mozilla (the “original” opensource browser) is now “competing” against webkit (the now-de-facto opensource web framework).
It seems to have lost it’s own ground somewhere along the line.
But it has certainly helped fantastically in shape what we have from the browser today.
But the age of the browser is still young – How can we make the interface more enjoyable to use & navigate? What new paradigms can we leverage? What new standards can we employ & develop? How can we take the browsing experience further?
These are questions all browsers are now starting to ask, & it’s brilliant to see.
Thanks for the update Mozilla. Let’s see if you can’t lose some weight now & get strongly back in the fight.

Comment by ProPuke — June 24, 2010

Good points giank and ProPuke. Maybe we’ll end up with an equilibrium (a la Emacs vs. Vi) where Firefox is a bit bloated and slow, but can do anything (like Emacs), and Webkit is small and nimble, but sticks to its core competencies (like Vi).
I, like many others, use Emacs for hardcore programming, but Vim for quick changes to system files. I wonder if this is how people will use Firefox/Webkit in the future…

Comment by Skilldrick — June 24, 2010

I really wonder how much JägerMonkey and hardware acceleration will help push Firefox (4) forward again.

Comment by SilentLennie — June 24, 2010

Leave a comment

You must be logged in to post a comment.