Hopefully this will also make it possible to have separate instances of the Flash plugin for different sites, rather than a single instance for all sites. In addition to improving security through compartmentalization, having separate copies of Flash will allow Firefox to unload each one when finished with a site. Flash leaks a huge amount of memory, and currently many people continuously have some site open which uses it (often sites like gmail or github that use it for "utility" purposes), which means Firefox can never unload it, resulting in an ever-ballooning memory footprint that people often blame on Firefox.
(I say this as someone who doesn't use Flash personally, but everyone benefits when systems have more security, and I personally would like to see less unfounded complaints about memory usage distracting from the legitimate efforts to improve memory footprint and performance.)
I'm pretty sure Firefox has run Flash in a separate process for some time now (so when Flash crashes, you get a stripy rectangle saying "oops, a plugin crashed" instead of your entire browser disappearing). I guess the news here is the extra-restrictive sandbox environment.
I think they put development on hold (Nov 4 2011):
"On Nov. 4, 2011, we held a public call to evaluate options for improving Firefox responsiveness including the multi-process Firefox initiative (code name Electrolysis, also known as e10s). The outcome of this discussion was a decision to put the Electrolysis initiative on hold"
I believe they got the plugins seperated out, but put a hold on going further and having individual tabs (or content vs chrome) running in different processes.
Yes. On Linux, where Adobe's Flash was piece of shit (sorry for the language but you can't describe it better), Firefox was using NSPluginWrapper for quite a some time. Similarly, Opera uses its own plugin wrapper. Both of these start Flash (and other plugins) in separate process so when the Flash crashes, it does not take down the browser.
Because NPAPI (the plugin API) is specified in terms of native OS handles (HWNDs on Windows, X ids on X), even a sandboxed plugin naturally has access to stuff like your keyboard, mouse, and can send keystrokes to your desktop. This sandboxing of Flash is a step but I don't think it will help much.
Flash in gone from my computers. I think I uninstalled it last Aug, but almost immediately had to reinstall for the online Stanford AI class, but I removed it after the final.
Flash will be around for a few more years, and it certainly was needed for the past several years. However, as a sort of "dog fooding", I would encourage developers to uninstall it so we can help build a better "HTML5" web quicker. YouTube has html5 support, for example.
In short, a large number of people exercising the HTML5 web, and addressing its shortcomings, would definitely help accelerate the move to HTML5.
While this solution obviously won't work for most users, personally I just reach for youtube-dl (http://rg3.github.com/youtube-dl/) when I run into a video that YouTube won't serve with the HTML5 player. No flash, no ads, and no buffering.
Safari's ClickToFlash/ClickToPlugin is absolutely awesome as it replaces the basic placeholder with a (apparently QuickTime X) player with an overlay offering source format selection (including Flash object, and a download option). Chrome has a similar plugin, although less advanced (and generally more flaky). Firefox's FlashBlock is so limited and looks like a stab in the eye in comparison.
Gnash--the free flash movie player--plays those YouTube videos fine. Except it doesn't show the ads :) It doesn't really work for almost anything else expecting flash though.
Mostly true, with the exception of Safari on OS X. If you have Flash installed on a Mac, Safari runs it as an isolated separate 64-bit process. This was not built by Adobe though, it was built by Apple.
What Mozilla also needs to address is the fact that Firefox update is now totally broken on a "normal" user account in windows, it will not even inform you that you are out of date. Before they changed to the "rolling release" model, it worked fine, hasnt worked since 4.0... So now there is going to be a non-insignificant number of windows users stuck on firefox 5 6 7 ... forever.
Is that number really "non-insignificant"? Most Windows users I know are always logged into administrator accounts. I myself use the administrator account on my home PC, even though I should know better. Well, at least I pay attention to UAC dialogs.
I actually like how Firefox updates itself. On a multi-user PC, updating Firefox updates every user's Firefox. Not so with Chrome.
This seems like a good move but they have a much greater need to copy Chrome's auto-update for Flash, both to deal with the massive legacy install base and to make any successful escalations from the sandbox less long-lived
Just in time for it to become more and more irrelevant. I say this with an amount of snark that may be frowned upon, but Flash has very few legitimate uses remaining. Though it's an enormous hack (that HLS will fix), I have live transcoded video streaming to an HTML5 video tag and I just got a minimal Skype clone working natively in Chrome. I rarely use Flash and the cases where it's required are quickly diminishing.
Codecs are still a problem, though - there's no single universal codec supported by all major browsers. WebM isn't supported by IE, Safari and iOS, H.264 isn't supported by Firefox, Opera and possibly Chrome in the future.
It's a mess, and I don't see it getting better in the near future.
Right now, two codecs cover all the browsers that matter. Internet Explorer supports WebM, but doesn't ship with the codec; Google provides the WebM codec for Internet Explorer (https://tools.google.com/dlpage/webmmf). That just leaves Safari and iOS that still only support H.264.
Meanwhile, if Adobe would ever get around to adding WebM in Flash as previously promised, that would work as a fallback for Safari, as well as IE users without a WebM codec, leaving just iOS.
Safari (on Macs) has had a codec plugin support mechanism for codecs for a long time via Quicktime, and I believe WebM is already supported in Perian, the popular ffmpeg/libav wrapper for OS X.
Interesting, but that doesn't seem like something straightforward to convince a random site visitor to install. Google's WebM plugin for IE seems like a hard enough sell even with Google's name attached to it.
I doubt it's beyond their ability, particularly as someone's demonstrated that it can be made to work. I'm guessing though that they'd be generally happier if people on Mac OS X just used Chrome. There's less likely to be Safari users who are forced to use the browser due to IT department dictates or whatever other odd reason people use to justify IE.
(I say this as someone who doesn't use Flash personally, but everyone benefits when systems have more security, and I personally would like to see less unfounded complaints about memory usage distracting from the legitimate efforts to improve memory footprint and performance.)