Firefox 3.6.4 Released With Out-of-Process Plugins

DragonHawk writes "Mozilla Firefox 3.6.4 went to general release today. The big new feature in this release is out-of-process plugins (OOPP). This means things like Flash, Java, QuickTime, etc., all run in separate processes, so when Flash decides to crash, it won't take your browser out with it. If Flash starts consuming all the CPU it can find, you can kill it without nuking your browser session. I've been using this feature since it was in the 'nightly build' stage, and it was still more stable than 3.6.3, just because Flash was isolated." And reader Trailrunner7 supplies another compelling reason to download 3.6.4: "Security researcher Michal Zalewski has identified a problem with the way Firefox handles links that are opened in a new browser window or tab, enabling attackers to inject arbitrary code into the new window or tab while still keeping a deceptive URL in the browser's address bar. The vulnerability, which Mozilla has fixed in version 3.6.4, has the effect of tricking users into thinking that they're visiting a legitimate site while instead sending arbitrary attacker-controlled code to their browsers."

Comments

First

[Shikaku]

Firefox post. Firefox is the fastest browser around!

Re:First

[shadowbearer]

I've been using Opera, Google's Chrome, and IE alongside Firefox on W7 for about four months now on three computers, on a consistent basis, meaning every day.

  Opera is a bit faster, Chrome is a lot faster, but we are talking about tenths of a second here when rendering anything other than extremely complicated web pages which to be honest would render a lot faster in any browser if the designers wouldn't include so much crap in them that demands connections to multiple websites for stupid things like a small advertising gif image from a server that is already overloaded.

  Over that time, Firefox has been easily the most stable browser I've ever used - that might have something to do with me running addons such as adblock, flashblock, and NoScript - denying access to a lot of the poorly written or implemented crap websites that can crash any browser. I can count the number of times that Firefox has crashed on all three of my computers on one hand since the beginning of the year - that's two laptops and one desktop, running combinations of Windows XP, Windows 7, Ubuntu and Fedora.

  It didn't used to be that way, no. But it is now. Firefox also consistently recalls my previous browsing sessions - even after the multiple downtimes I had tonight during numerous power outages due to bad storms (the new battery for the UPS is in transit and should arrive tomorrow, and I ordered it from a website that does not list Firefox in their supported browsers list) neither Opera nor Chrome did so.

  The addon Xmarks has proven to be both useful and consistently stable, I'd highly recommend it.

  YMMV, YEMV, etc. This is just mine. I don't know about the rest of you, but I'll take stable over fast any day. I regularly have from a dozen to several dozen tabs open at any one time, and being able to recover my work after any crash, no matter the cause, means a lot to me. These features should have been written into browsers as DEFAULT features from the beginning. Somewhat around ten years ago I remember wishing that someone would just code a browser that could remember what I was doing before a crash, and do so consistently. Now, finally, I have one. Thank you, Mozilla.

  What I find ironic about the whole browser war is that the "feature leader" over the last decade has been the open source solutions - specifically firefox, and the rest of the field is playing catchup - especially Microsoft.

  SB

 

Re:First

[KingMotley]

I would agree with your assessments in Opera and Chrome. I'd really like to like Chrome, but it's still missing firebug, and that unfortunately is pretty important to me.

However, your experiences with firefox stability and mine are completely different. You say you can count the number of times firefox has crashed since the beginning of the year on one hand, while I can barely count on one hand the number of times it's crashed on me TODAY. I am hoping this release reduces that number significantly.

Re:First

[shadowbearer]

  See my other post regarding background programs, in particular antivirus programs. I have only run Avast here for many years, and it's one of the few common factors between my machines and my customer's machines, where Firefox seems to work just fine.

  I don't have and probably never will have enough data to know for sure, but I suspect that antivirus and some malware scanners might contribute to FF stability somewhat. Other than that it's kind of a crapshoot.

  However I'd bet money that if your machine is hardware/driver stable, I could build it to be software stable. Made it so for hundreds of people over the last couple years. Post back and if you're willing to share some email I can try and help.

  Gotta crash myself, tho, been up too long just like Win98 ;-)

SB

Re:First

[ultranova]

Opera is a bit faster, Chrome is a lot faster, but we are talking about tenths of a second here when rendering anything other than extremely complicated web pages which to be honest would render a lot faster in any browser if the designers wouldn't include so much crap in them that demands connections to multiple websites for stupid things like a small advertising gif image from a server that is already overloaded.

The real problem with these pages is not that they are slow to render, but that the rendering sometimes has to wait for that advertisement, and you see nothing until it's loaded. Even more annoyingly, the whole browser UI sometimes freeses to wait for it.

I wonder if we'll see less of this now that plugins are separate?

Re:First

[nmg196]

> Firefox is the fastest browser around!

Apart from Chrome and Internet Exporer 9 :)

Re:First

[imakemusic]

Ah! Internet Explorer, The Ayrton Senna of internet browsers - It's got the speed, but somehow that doesn't help.

Re:First

[Spliffster]

Did anybody hear that **WHOOOOOSH**-ing sound ?

just asking.

-S

Re:First

[cgomezr]

I'm afraid Firefox hasn't been the feature leader at all. Tabbed browsing? Opera had it before. Mouse gestures? Opera had it before. Quick dial? Opera had it before. Customisable search bars? Opera had them before. Ad blocking? Opera had it before (although, admittedly, worse than Firefox's). Stored sessions? Opera had them before (and it does restore from crashes without any problem in my case). I could keep enumerating, I'd say 90% of the browser features that Firefox implements are copied from Opera.

OK, I think Firefox had private browsing before Opera, making it the browser of choice for pr0n (i.e. 99% of the internet usage); but now Opera has catched up on that and offers private and non-private tabs mixed in the same window :)

BTW, on my machine Opera behaves much better than Firefox with 20+ tabs open (I have 57 right now), it's still snappy and Firefox would be crawling and taking up loads of RAM. But of course YMMV.

Re:First

[Anonymous Coward]

I'm sorry, but I use Opera about 90% of the time and Firefox (with noscript) the other 10% of the time, and I still experience about 3 times as many Firefox crashes, and Firefox tends to lose my previous browser sessions far more often, especially after a crash.

Additionally, I wouldn't classify Opera as "a bit faster". It's leaps and bounds faster. I'm not talking about rendering time, however, since I've never really noticed a difference there in any non-IE browser. I'm talking specifically about loading times (firefox always takes about 5-10 seconds to open, about 15 seconds the first time it loads after a reboot, while Opera is about 1-3 seconds), responsiveness when scrolling, switching between tabs, etc.

I love Firefox for the firebug plugin and other developer tools, it's so much better than Opera's Dragonfly in almost every way. Without Firefox my web development times would probably increase by about 30%. I can't imagine using Firefox as my primary browser though, I'd probably switch to Chrome, Safari, or give up using the internet altogether ;)

Re:First

[Anonymous Coward]

So, you do realize you said FireFox is only a tenth of a second slower than Chrome. .. because you're not downloading JavaScript, Flash etc.

Chrome positively SCREAMS on each and every machine I've tried it with.
FireFox can go fuck itself. And so can you.

Re:First

[nstlgc]

I don't think "ironic" means what you think it does.

Re:First

[Canazza]

Chrome has built in dev tools that mimic pretty much what firebug does. It even does breakpoints in javascript. It's pretty much missing proper adblocking and flashblocking (I'm using adblock for chrome and it's one unwieldy bastard tbh). otherwise it's a good browser.

I still prefer firefox though.

Re:First

[vbraga]

That made my first laugh today in a very stressful week :) Thank you, sir!

Re:First

[Jurily]

The main problem I see with web browsers today, is that they completely and utterly ignore every single user interface design convention they can find.

With Chrome reinventing window layout, Firefox reinventing standard dialog layout, and Opera reinventing UI themes, where do we take refuge? Hell, even IE doesn't have menus by default anymore.

That said, Firefox has Adblock, and Adblock has hufilter, so I'm not switching anytime soon.

Re:First

[Anonymous Coward]

But NOBODY CARES ABOUT OPERA!

Re:First

[Anonymous Coward]

lol opera? r u serial?

Re:First

[sandertje]

FF might save your data in a crash, but in doing so it gets tantalizingly *slow*. I usually have 10+ tabs open at any given time, but on the next restart, when FF restarts and tries to open all tabs at once it just slows down and down, quite possibly clogging up the machine in severe cases. The end result: most of the time I'm just hoping for the "well, this is embarrassing" message ;-).

Re:First

[Pax681]

classic one sir!!!!! genuine belly laghs from that line!!!!

Re:First

[shadowbearer]

  True; my last comment was aimed at IE fanboys, I'll admit ;-)

  YMMV, true. I do like Opera, it is my second browser of choice. I still find Firefox more stable and I like the fact I can remove features I don't want.

SB

Re:First

[Anonymous Coward]

You know you're doing it wrong when you have 57 tabs open

Re:First

[icebraining]

Use NoScript and all the plugins (not only Flash) are blocked by default, along with external scripts.

Re:First

[Anonymous Coward]

I also use the latest versions of all the 'big' browsers daily for work and Firefox is definitely the least problematic, most stable and feature rich - that's why it's installed on all my personal boxes.

With Firefox, it's quite usual for me to have 120+ tabs open, all restoring properly between sessions (I'm lazy and don't like searching again when FF just keeps it all under control). I leave all the 'other browsers' for testing purposes only.

Just my experience.

Re:First

[Runaway1956]

Yeah - FF does get rather bogged down from time to time. I'll admit though, for myself, I suspect a lot of that bogging is due to the user scripts I have installed. I play an online game, and one of my scripts does my scouting for me. During a long scouting run, that and all other Firefox windows get so sluggish, I'm just waiting for it to all come crashing down.

The scripts have been updated recently, and the bog isn't so bad now - but it's still there.

People who don't install bandwidth and memory hog scripts may never notice this problem.

Re:First

[Maxo-Texas]

I downloaded this and flash 10.1 last night.

Farmville (which had become unplayable and killed my browser so bad I was using a second computer) became playable again.

So I'm happy- I suppose. May have to quit farmville soon because of work tho. Lots of very intense weeks and months coming up.

Re:First

[nabsltd]

BTW, on my machine Opera behaves much better than Firefox with 20+ tabs open (I have 57 right now), it's still snappy and Firefox would be crawling and taking up loads of RAM. But of course YMMV.

I'm wondering how much other apps/OS settings have to do with this.

On my work machine, I see FF taking up to 1.3GB of RAM, and it becomes very unresponsive until I close and re-open the browser. On my home machine, it never takes more than about 300MB, despite essentially the same browsing pattern.

The work machine runs XP SP3 while at home I'm running XP 64-bit. I run the 32-bit version of Firefox on both.

Re:First

[Deagol]

FreeBSD/amd64 8.1-PRERELEASE with firefox-3.6.3,1. Uptime 2.75 days, with the browsing being open almost as long without shutdown. At this moment, I have 11 tabs split across 2 windows (one fullscreened on 2 different virtual desktops). Virtual: 1252M. Resident: 1100M. It starts fresh at between 200 and 400MB, then continues to climb. By the end of the week, it'll be pushing 2GB resident. Glad I have 8GB to throw around.

Plugins: WikiLook, Web Developer, User Agent Switcher, Screengrab, Noscript, Linkchecker, HTTPS Everywhere, Ghostery, Font Finder, Firebug, Download Helper, Adblock Plus, and Customize Google.

Re:First

[Anonymous Coward]

"BTW, on my machine Opera behaves much better than Firefox with 20+ tabs open (I have 57 right now)"

You don't need a browser, you need ADD medication.

Re:First

[kbrosnan]

If you are on Windows or Mac OS X try typing about:crashes in the address bar should provide some crash reports. If you would like reply to this with the ids and I'll take a look.

Re:First

[Zancarius]

BTW, on my machine Opera behaves much better than Firefox with 20+ tabs open (I have 57 right now), it's still snappy and Firefox would be crawling and taking up loads of RAM. But of course YMMV.

Opera's UI doesn't handle >200 tabs gracefully at all, regardless of how much better behaved it is. In fact, I have yet to find a browser besides Firefox that handles multiple hundreds of tabs in a manner that is relatively easy to navigate (I take that back, I think Arora behaves very similarly to Firefox). That's not to say Firefox doesn't have its issues: Since 3.0 was released, I haven't been able to keep > 300 tabs up without Firefox crashing fairly regularly whereas the 2.x branch was able to maintain about 500 tabs fairly well (if sluggishly) and rarely ever crashed.

Now, I know what you're going to say: "more than 100 tabs? That's crazy and pointless. Don't do it."

Yeah, well, it's my particular use case, and judging by previous comments on /., I know I'm not alone because I know of one particular poster who ousted me with more than 800 tabs on a regular basis. Since I know someone is going to tell me that there's absolutely no reason to have more than 20 tabs open at a time, I'll explain: It's easy. People like myself and the other guy who had an average of 2 times more tabs open than me are the sort who middle-click on any story or item that might be vaguely interesting, and we therefore use open tabs as sort of an active history of information we're in the process of digesting. My last 300+ tab session I closed through had probably 40 tabs open to Slashdot stories over the course of about 2 months, at least 60 Wikipedia articles I'd opened over a similar period, various news articles, at least 20 abandoned Google searches. I also had ~100 tabs opened to various pages of documentation/manual mirrors hosted locally that I'd largely forgotten about. The other 80ish tabs were a mix of stories and other miscellaneous links I had followed. I find that JS-heavy sites like Gmail tend to kill Firefox if it's been running for about a day with 250-300 tabs open, so I don't hold out a great deal of hope that out-of-process support for Flash will completely fix the problem. If nothing else, I can at least go back to viewing Youtube; though, I tend to do that in Chrome or Opera simply because they do lack the sluggishness and jerkiness that is becoming of viewing online video in Firefox when it's under undue stress.

For anyone who doesn't browse in this fashion, you won't understand why it's so appealing, but I find it more efficient for my particular use case. My memory for specific sites isn't all that great, but my memory for recalling approximate locations/proximities of tabs and items of interest is pretty good. So, I let the application do what applications do best: I let it remember all the details for me, and then simply return to what I was doing before with regards to reading or researching. Besides, bookmarks don't really cut it for me, and I'm sure we've all been there: Dozens of bookmarked sites and not a damn clue where you'd found that one thing that was really neat. So, you troll your browsing history for a half hour and finally relent that you should have simply Googled it in the first place! Really, bookmarks (and even Google Bookmarks to a small degree) always seem to do this: They bring me straight back to having to use a search engine or other search utility (yay, grep!) to find something I may have accidentally closed a meager two hours previous, and if I didn't browse in the rather odd fashion that I do, I'd imagine I'd have to resort to searching a lot more frequently. As it stands, I generally only close my plethora of tabs whenever a) Firefox becomes too unstable to continue (I hope this gets fixed) and/or b) I'm fairly certain those tabs will no longer have a great deal of utility in the near future. Once they degrade to B status and haven't been bookmarked, I bookmark them.

I li

Re:First

[wallsg]

I'm afraid Firefox hasn't been the feature leader at all. Tabbed browsing? Opera had it before. Mouse gestures? Opera had it before. Quick dial? Opera had it before. Customisable search bars? Opera had them before. Ad blocking? Opera had it before (although, admittedly, worse than Firefox's). Stored sessions? Opera had them before (and it does restore from crashes without any problem in my case). I could keep enumerating, I'd say 90% of the browser features that Firefox implements are copied from Opera.

It would appear that Mozilla has learned from Microsoft to Embrace and Extend.

Re:First

[Anonymous Coward]

I have been running a somewhat large javascript program on Firefox, Opera, and Chrome for the last several months. Chrome has bee *much, much* faster. Firefox and Opera - about the same.

Re:First

[Anonymous Coward]

Wow, jack off to Opera more, jesus.

Great

[Darkness404]

This is great, now if only Firefox could separate tabs into processes and get a JavaScript engine comparable to V8 they could start to pull ahead of Chrome.

Re:Great

[BhaKi]

For performance reasons, tabs don't and shouldn't run in separate processes. You know, the original motivation for the tabs feature was that each tab could be run in a separate thread whereas each window needs a separate process. On most platforms, processes are more expensive than threads.

Re:Great

[BZ]

> For performance reasons, tabs don't and shouldn't run in separate processes.

In both IE8 and Chrome, tabs do in fact run in separate processes, with some caveats.

> On most platforms, processes are more expensive than threads

While true, processes have the benefits that:

1. One process crashing doesn't bring down other processes.
2. A process can be run in a low-privilege mode and memory isolation keeps it from accessing the guts of other processes.

Threads don't have those two properties, and both IE8 and Chrome actually make use of them.

Re:Great

[jrumney]

On most platforms, processes are more expensive than threads.

True, but on modern platforms the difference is not as significant as it was in the days of Windows 95.

Re:Great

[keeperofdakeys]

I'm really hoping that separating the tabs will help with memory leaks. Although there is a cost of memory using separate processes, the leaked memory will close with the tab and not stay for the entire session.

Re:Great

[exomondo]

For performance reasons, tabs don't and shouldn't run in separate processes.

What performance reasons?

On most platforms, processes are more expensive than threads.

That's not a valid reason to suggest threads are more ideal than processes. In fact given the nature of a multi-tab web browser - where the bulk of the processing functionality is duplicated - processes are more suitable, at that level of abstraction you pretty much have one browser per-page anyway, just all sitting in one window. Sure things like inter-thread communication are faster than inter-process communication but in this scenario you don't need that communication so there is no benefit there.

I would suggest in this case you are better off with a process-per-page model than a multi-threaded renderer since - as was said earlier - the scheduling is handled by the OS. One single-threaded (or at least single-page, you'll get benefit out of spawning workers for some tasks) renderer per-page is going to be more efficient than one renderer handling scheduling and spawning worker threads to render multiple pages as it has to manage all that synchronisation as well.

Re:Great

[exomondo]

Excuse the self-reply but what i meant by:

the scheduling is handled by the OS

should actually include synchronisation and management, insofar as once you've spawned the new process it takes care of itself, you don't need to worry about cleaning it up (though you can kill it) and you don't need to worry about thread-management.

Re:Great

[pinkushun]

Don't forget that Firefox was The One Browser that changed the browser-scape, with proper HTML and CSS implementations, well... needless to say it was miles ahead of others, for a long while. Remember our progenitor, make obeisance.

Re:Great

[Zoidbot]

Surely Opera is now the benchmark for performance, not V8. (10.60 beats V8, even in Chrome6 by some margin). Opera is also more complain with Javascript according to Googles own Sputnik tests, it's also has better HTML5 and CSS3 support...

Re:Great

[Anachragnome]

"On most platforms, processes are more expensive than threads."

To be honest, I don't think it really even matters.

While having this very page open in one process, I just opened 20 more Firefox processes, all loading the Slashdot main page. I simply clicked my shortcut for it 20 times in rapid succession, and they were all opened and loaded within 3 seconds of my last click.

I then closed them all (except this one), then quickly did the same thing, but opened 20 news tabs (in this process!) to the same page. They were all opened and loaded within 3 seconds of my last click.

Doing both of these things never resulted in more then a 2% increase in RAM usage and CPU usage dropped to less then 6% when the last process/thread opened, all on a P4 Processor with Win Xp and 2GB RAM. No crashes, no stutters, everything worked as intended both methods.

If I can't tell the difference at that point, what does any of this matter?

It's like the difference between 60fps in a video game and 70fps. Either is more then good enough.

Re:Great

[kangsterizer]

However processes use a lot more memory. Firefox uses way, way less memory than Chrome when you have a few tabs open.

Also, the browser should not crash. But if it does, it restore the session, but seriously, that rarely happens on Firefox (yeah, Chrome tabs crash all the time, but that's Chrome's fault... flamebait maybe but one could argue tab-process encourage buggy code since it's no big deal when a tab crashes)

The only things the browser does not have control over are plugins, and they're not in their own process, which is cool. Extensions are a more complex matter, I suppose they could still bring down everything with own process tabs.

I'm not sure the security added by sandboxing tabs into processes is worth the trouble right now. It's some kind of hack after all.

Re:Great

[Dan Ost]

At least on Linux, the memory footprint of spawned processes is minimal since the memory space is treated as copy-on-write.

Because of this, there is no significant difference in the amount of memory used to spawn a thread or a process.

Re:Great

[Darkness404]

Ok, how much memory does the average system have? A lot. Does it -matter- if Firefox uses 2 GB of memory if I have 4+ gigs? Memory is dirt cheap now, even insanely cheap systems come with 2 GB standard. Heck, phones are starting to get 512 MB + of RAM! Plus, memory is rewritable, reusable and if you aren't using it it won't be utilized. Would you rather gain a performance benefit by using 2 GB of RAM if you've got no other major programs running on a system with 4 gigs of RAM, or rather have a few GB of idle RAM to lower memory usage? Low memory systems now have at least 1 gig of RAM, embedded systems now have 256 MB + of RAM. Memory is dirt cheap, might as well use it.

Re:Great

[BZ]

> However processes use a lot more memory

Depending on how you measure, they can, yes.

> Also, the browser should not crash

Sure would be nice! ;)

> I'm not sure the security added by sandboxing tabs into processes is worth the trouble
> right now.

Why not? It's as effective an exploit-mitigation strategy as anything else out there. Reduces attack surface, etc...

Re:Great

[Malc]

If you're worrying about memory, then per-page processes is actually a boon. If implemented properly, one should be able to figure out problematic pages and close them, re-open them. FF is a total memory sieve these days, so I'm really hoping for something like this in FF4. It's pain at the moment having close the whole browser and then wait for 40 tabs to reload.

Re:Great

[marcosdumay]

Linux processes are created with a 4kB stack by default, Windows processes are somewhat bigger anything else either can be shared or is specific for a single tab. Anyway, I doubt it is that few kB of extra memory that you is complaining about, so I'm having a hard time understand how processes use a lot of memory. Ok, on a web server with 10k processes running, that is important, but for only some 100 or so open pages? That adds to 400kB of RAM.

Re:Great

[keeperofdakeys]

It's not the memory of spawning new processes, but the fact that you need to have the browser runtime in the memory of each of the new processes.

Re:Great

[smellotron]

Does it -matter- if Firefox uses 2 GB of memory if I have 4+ gigs?

Yes it does. My computer is more than a web browser kiosk. It only takes two applications with your attitude to force the OS to start swapping.

Re:Great

[Dan Ost]

What do you mean?

The executable code will be stored exactly once no matter how many processes show it. The data that is changed (and therefore copied) per process would be exactly the same data that is changed per thread, so the memory requirements should be near identical (not quite identical since memory is copied at the page level, but threads could share pages).

What am I missing?

Re:Great

[adrianmonk]

I'm really much more interested in what the two browsers do when you have 25+ tabs open at one time. That's the case that's likely to have consequences for system-wide memory usage. Even more than that, I'm interested in the case where you've had the browser open for 8 days and have opened and closed hundreds of tabs and thousands of pages.

Having separate processes buys the ability to eventually close a process (when all the tabs/pages associated with it are closed) and reclaim all the resources. This is handy if there are memory leaks (with proper programming, there shouldn't be, but there *certainly* have been in Firefox -- a browser is a pretty complex piece of software). It also seems like it would be helpful for memory fragmentation: if you keep the entire browser state separated into N buckets (processes' memory spaces), then when you free one bucket, you can free every single page in it. You never find yourself in a situation where you can't free a page because some unrelated thing (that happened to call the same allocator you did) has 100 bytes of data still occupying that page.

You're right that it could lower the consequences for sloppy programming, but I don't care that much. Anti-lock brakes on my car lower the consequences for sloppy driving, but I still want them.

UI Lag

[electrosoccertux]

now can we do something about the rest of the awful browser?

Open 20 tabs and the entire thing chugs to a grinding halt as only one (1) of my four (4) processor cores gets maxed out. So much for the "multithreading" everybody says that Firefox.
The same list of 20 tabs peg all my cores to 100% for a few seconds and then they're all done rendering, when I'm using Chrome. No thanks Firefox. You guys are ancientsauce.

Re:UI Lag

[Nadaka]

I have never had problems with firefox having a ton of tabs open.

I regularly have 15+ tabs, sometimes 50 or 60. The only time I have any issues is if I turn off no script and get some flash or javascript running to slow things down.

Re:UI Lag

[Darkness404]

So in other words, the thing runs perfectly if you disable the default options and install ad-ons to make it work right and then disable plugins.

Lets face it, a good browser wouldn't require those things to be done in order to browser decently.

Now, I use Firefox, but only because I hate the lack of customization on Chrome, don't like the proprietary-ness of Opera, run Linux so can't really use Safari, and obscure WebKit/Gecko browsers usually don't have needed plugins like AdBlock and manytimes don't have enough customization.

Is it too much to ask for Chrome's rendering engine with Firefox's UI only if it was a bit faster?

Re:UI Lag

[dakameleon]

Don't forget the ponies!

Re:UI Lag

[Anonymous Coward]

My problem is that I kind of hate the Internet with NoScript. I just gave up on whitelisting nearly everything and instead started trying to blacklist the more egregious ads.

In my experience on my hardware (various machines, so it isn't just a one-off), none of the other major browsers -- Chrome, IE, Safari, Opera -- seem to have nearly so many UI mini-freezes that happen nearly constantly for me on Firefox, which is probably the biggest single reason I eventually dropped it as my default browser. I don't mean this to start a browser pissing contest -- I can name things I hate about every one of those browsers, and I'm sure there's a cadre of people who have never seen my issue on Firefox but have seen it on the others, but this is the thing *I* hate about Firefox and I have collected a few anecdotes of others with the same experience.

If it's really flash, then this sort of isolation can go a long way toward mitigating it; but if it's javascript or something else, then I really hope Mozilla invests some time addressing the responsiveness problems.

Re:UI Lag

[Anonymous Coward]

No idea what you're doing to get slowness with 20 tabs. Does your machine have a memory bottleneck perhaps?

With 128-150 tabs open, I had to shut down about once every one or two days. The more flash-based sites I had open, the more quickly it would slow down and/or crash. Sometimes javascript-intensive sites would also be problematic, though I have many more of those open. And again, it tended to last longer than 24 hours before a restart was needed.

That was before this update. I'm curious to see how long 3.6.4 will last, given that the plugins are now isolated. The UI seems a little snappier, at least.

Now 200+ tabs: THAT really causes a slow down. In my experience, the difference between even 190 tabs and 210 is quite dramatic. They should really look into that.

Re:UI Lag

[Fnkmaster]

Chrome's rendering engine is slow and sucks. Javascript is much faster, but that's it.

Re:UI Lag

[nmb3000]

This, this, this, this, this. The terrible user interface responsiveness of Firefox is what kept me on IE for the longest time (and I only moved because of addons, not because Firefox itself is any better).

For a good test, open a Slashdot story with ~1000 comments and watch as the browser just stops dead in the water for 5-15 seconds while it renders the page. You can also try opening the browser when you have 10 or more tabs saved in your session. Again, the entire interface is useless while the pages are rendering. If the browser really is multithreaded in any meaningful fashion, then the rendering threads obviously have a priority higher than the UI, which seems like a bad thing.

I'd rather have this improved than move plugins into an external process. Since I started using NoScript I haven't had Firefox crash because of Flash. Ever. However, I still read Slashdot so I do deal with the lagging on a regular basis.

Re:UI Lag

[nmb3000]

With 128-150 tabs open

No offense, but I think you're doing it wrong.

Re:UI Lag

[nmb3000]

My problem is that I kind of hate the Internet with NoScript.

You might take a look at YesScript (a JavaScript blacklisting plugin sans all the extra protection crud in NoScript). If you use it in conjunction with AdBlock+subscriptions you'll probably block quite a bit.

That said, I like NoScript in general because of just how much faster most sites are with their scripts disabled. It does get annoying though, as more and more sites are completely non-navigable without scripts enabled.

Re:UI Lag

[Ethanol-fueled]

I'm sure there's a cadre of people who have never seen my issue on Firefox

Yup.

but if it's javascript or something else, then I really hope Mozilla invests some time addressing the responsiveness problems.

Yeah, we get it, paid shill of Google. Chrome runs Javascript 1 millisecond faster than Firefox does. Perhaps you should upgrade from that Pentium II 233 and run an O.S. that dosen't have 1-minute rendering overhead from AV software and other crap.

Then come back.

Re:UI Lag

[Anonymous Coward]

Tabs are the new bookmarks.

Re:UI Lag

[hairyfeet]

Are you on Windows or Linux? Because I've noticed the Linux version just doesn't run as well as the Windows version. If you are on Windows may I suggest you run PCWizard by the guys at CPU-Z, as it sounds like you may have a bottleneck causing problems. I've seen weird slowdowns in things like browsers before, and using the benchmarks in PCWizard have been able to trace down the culprit.

Because my PCs aren't top o' the line by any means (an AMD 925 quad with 8Gb and a 1.8Ghz Sempron single with 1.5Gb) and since going to the 3.x branch FF has been nothing but zippy, which is pretty impressive considering how many tabs and extensions I have, plus the fact I only use sleep on my PCs, never reboot except for updates. Currently I have ABP, NoScript, Downloadhelper, Downloadstatusbar, FEBE (a must have IMHO) ForecastFox, iMacros (another must have) Imagezoom, nightly tester tools, and distrust. And neither the Sempy nor the quad has been anything but snappy in FF.

So run PCWizard (there is a zip version so you don't even have to install it) and run the individual benches. Then look at how you compare to others, and if something is way below what you have then something is up. I had the same thing happen to me on a dual Athlon and it turned out a shitty chipset driver was causing all sorts of little slowdowns.

Re:UI Lag

[kiddygrinder]

chrome's got an alright engine, they just need to fix up the interface... sticking to firefox for now.

Re:UI Lag

[Anonymous Coward]

Javascript execution time is not the same thing as UI responsiveness. Firefox could have a javascript engine 5 times as good as Chrome's, hypothetically, and still block UI on javascript execution. Or it might be something other than javascript. The GGP brought up NoScript, which is how Javascript came up.

Don't cry shill and then invent a straw arguments. I don't even get why you assumed that he was shilling for Google.

Be less of an asshole and learn to think critically. Then come back.

Re:UI Lag

[compro01]

For a good test, open a Slashdot story with ~1000 comments and watch as the browser just stops dead in the water for 5-15 seconds while it renders the page

I haven't found a browser that doesn't do that. Firefox, IE8, Chrome, Safari, and Opera all do that for me, at least on Windows. Haven't done any meaningful testing on Linux lately though.

Re:UI Lag

[Ethanol-fueled]

Be less of an asshole and learn to think critically. Then come back.

Quoth my parent, posting anonymously:

f it's really flash, then this sort of isolation can go a long way toward mitigating it; but if it's javascript or something else, then I really hope Mozilla invests some time addressing the responsiveness problems.

Like many other posts all over the discussion, that one implicitly(and other posts more explicitly) touts Chrome's Javascript speed advantage, as if they were parroting other articles and putting all of one's eggs in one basket to shill the trivial performance gain to generate some Chrome hype. Or perhaps I'm a bit slow, and the average Slashdot reader has a 1-millisecond resolution and is sensitive to things I'm not.

And my hardware is a 6 year-old laptop.

Re:UI Lag

[Shakrai]

Since I started using NoScript I haven't had Firefox crash because of Flash. Ever.

I have, but it was always traced back to a corrupted profile. Deleting the profile and starting over has always fixed it for me.

Now if I could only figure out why 3.6.3 keeps freezing on me for no apparent reason. 3.5.9 never gave me any issues with the same configuration and add-ons but 3.6.3 freezes up at least one a day. Interestingly enough when it freezes it doesn't max out the CPU -- just sits there without responding at 0% cpu usage.

Re:UI Lag

[RobertM1968]

now can we do something about the rest of the awful browser?

Open 20 tabs and the entire thing chugs to a grinding halt as only one (1) of my four (4) processor cores gets maxed out. So much for the "multithreading" everybody says that Firefox. The same list of 20 tabs peg all my cores to 100% for a few seconds and then they're all done rendering, when I'm using Chrome. No thanks Firefox. You guys are ancientsauce.

I am curious if that is a Windows specific problem (not as in "MS screwed up" but as in "Firefox for Windows does not take advantage of SMP") as I have no such problem on eComstation or OS/2 Warp. Any Linux users who can confirm this problem does/does not exist for Linux?

Re:UI Lag

[Anonymous Coward]

Actually I have the same "all of ff pauses while it loads the page" problem, If I try clicking around firefox turn white and I get a busy cursor. I can't even change tabs
You can pretend that everyone that has problems with firefox is a shill but the reality is that for some reason, it does not work the same for everyone.

If you really want to advocate for FF, perhaps helping people figure out what the problem is would be better than denying it and calling them names.

If I'm lucky, this new version will fix it, I'm still going to wait a few days to make sure that there's no gotchas in the new version.

Re:UI Lag

[Anonymous Coward]

Is it too much to ask for Chrome's rendering engine with Firefox's UI only if it was a bit faster?

Yes. The Firefox UI is implemented using an XML doc that describes the UI. Google for "XUL". Add-ons can override any part of the UI by overriding the XUL that describes the part of the UI they want to change. This design is extremely flexible, and allows experimentation and rapid change in the UI. However, it is not going to be as fast as chrome, because chrome is using the normal OS APIs directly to build its UI.

Re:UI Lag

[Anonymous Coward]

Whitelisting each domain *once* is hardly annoying unless you have many computers, in which case I'd be looking into sync'ing each machine to a common config.

I find the real relief in NoScript disabling all the domains/scripts that aren't necessary to the function of a page, such as (ad) partners and miscellaneous third parties.

Re:UI Lag

[Anonymous Coward]

I have never had problems with firefox having a ton of tabs open.

I regularly have 15+ tabs, sometimes 50 or 60. The only time I have any issues is if I turn off no script and get some flash or javascript running to slow things down.

Come back when you've had over 400 tabs. That's when you really start having probelms...

Re:UI Lag

[inode_buddha]

Interesting. Ever since my SMP box died, I'm using an old P4 e-Machine with 512 megs and linux. Flash playback, and video in general plays just fine. Graphics are onboard Intel i915. Though newer versions of FF *are* much better. I saved a bunch of CPU horsepower by using a decent hosts file so that AdBlock and NoScript don't have to work so much.

The UI *does* lag a bit with pages that have tons of comments, but not nearly as bad as it used to be. On the SMP box there wasn't any lag at all. By SMP I mean multi-sockets and large RAM; not just multi-cores.

Re:UI Lag

[slifox]

I keep firefox sessions open for months at a time, with 150+ tabs.

I use 64-bit firefox v3.5.9 on Debian linux, with 19 extensions, java disabled, and flash isolated using 32-bit flash + nspluginwrapper (meaning that flash runs in a separate process for compatibility, and the huge extra benefit is that flash crashing can't take down the browser).

Javascript is fully enabled, though I do have AdBlock to remove annoying ads.

I have no problems: firefox runs very fast (pages render very quickly) and it doesn't leak memory (though 150+ tabs DOES use 1GB+ memory, but I have 4GB total and lots of swap)

In short: Firefox performs remarkably well under extreme circumstances. I doubt Chrome or Safari could deal with these cases, and even if they could, they don't have the extensions that make them usable with so many tabs (mainly Tree Style Tabs -- best extension ever!).

Re:UI Lag

[Randle_Revar]

My experience is that it runs better on Linux than on Windows...

Also, what the heck happened to D2? It's like I'm back in the early Aughts.

Re:UI Lag

[Randle_Revar]

My setup is similar, although my main browser is SeaMonkey 2.0 nightlies, and I run NoScript. Although it takes some time to load that many tabs, it doesn't cause any problems.

Re:UI Lag

[Randle_Revar]

>My problem is that I kind of hate the Internet with NoScript.

really? I hate the internet *without* noscript

Re:UI Lag

[drsmithy]

Open 20 tabs and the entire thing chugs to a grinding halt as only one (1) of my four (4) processor cores gets maxed out. So much for the "multithreading" everybody says that Firefox.

I run 150+ open tabs in Firefox (times two, because I have a "work" and "personal" instance of Firefox Portable), all day, every day. Sure, it crashes every few days (Session Manager to the rescue), but calling 20 tabs anything significant, is laughable.

Re:UI Lag

[surveyork]

Firefox has known bugs regarding slowness when dealing with long pages. Some of these bugs have remained unsolved for years, perhaps because they mostly involve rare cases. Go to bugzilla.mozilla. Search for unresolved core bugs with "slow" in the summary. ???? Profit.

Re:UI Lag

[swilver]

20 tabs? Try 500+ tabs... with flashblock, noscript and a tab counter plugin (obviously). Yes with Firefox. Only time it annoys me is when the browser crashes (due to lack of memory, it doesn't like it when it comes close to 2 GB).

Re:UI Lag

[DigiShaman]

I noticed the exact same issue that nmb3000 is having regarding Slashdot with Firefox.

I'm running Win7 (64bit) with an Intel Q6600 (quad core), 8GB RAM, and nVidia 275 video card.

Re:UI Lag

[Anonymous Coward]

Last night I had every RFC open in a tab on a machine with 512MB/RAM (and no swap).

Sure, they're only ASCII, but that's nearly 6,000 tabs - on a nightly build.

Re:UI Lag

[shadowbearer]

So in other words, the thing runs perfectly if you disable the default options and install ad-ons to make it work right and then disable plugins.

  I'm running the release with over sixty tabs open, adblock, noscript, flashblock, + other addons, an HD youtube video for entertainment on the second monitor, several adobe plugin pdfs open, plus some active weather flash running (it was storming here earlier, watching the radar) and Firefox is only using about six hundred MB or so. My three year old desktop X2/32bitW7/4GB is still snappy, I hardly notice the difference.

  I don't even remember the last time Firefox crashed on this system (W7). Sometime in February I think, I'd have to look at my logs. Firefox has been incredibly stable for me for at least a couple years, and that experience has been echoed on the systems I build for customers as well. I suspect at least some it may be due the other memory resident programs on the computer, particularly antivirus programs, although I can't name any offhand, not enough data yet.

SB

Re:UI Lag

[Spad]

So in other words, the thing runs perfectly if you [...] install ad-ons to make it work right and then disable plugins.

...and obscure WebKit/Gecko browsers usually don't have needed plugins like AdBlock

Re:UI Lag

[fearlezz]

I'm having problems with firefox freezing as well. I always assumed it was my crappy video card configuration, but with FF 3.6.4, a lot of speed issues are actually fixed. They've done a great job on this release.

Re:UI Lag

[hack slash]

The biggest annoyance about FF I have is just that - why is FF giving CPU time to tabs you haven't looked at for a while?

It should be that you can configure something like "tab CPU timeout" in minutes so when you view a different tab, after X minutes the tab which is no longer displayed gets no CPU time at all - this should keep the browser fast even when you've got 10's/100's of tabs open.

I keep dozens of tabs open on my main machine as I use it as an alternative to keeping bookmarks, saves the hassle of clicking bookmarks and reloading whole pages - flipping to a different tab is like turning a page in a book, the information is there instantly, but it shouldn't suck CPU power when you're not looking at it.

Re:UI Lag

[l0b0]

Sorry, but you might want to check up on what web sites, extensions and spyware you're running. Computers have this annoying habit of doing what you tell them to do, rather than what you want them to do. I've been using Firefox since 0.6, and I keep seeing these comments about huge problems, never* experiencing them myself. I don't have high end machines, and I've never had more than 2 GB of RAM. Been running Firefox on Windows XP, then FreeBSD (keep up the good work!) and now Ubuntu.

* As in, whenever I do, it's trivial to identify the culprit /on the web site/ - Java applets (those were the days), Flash (hi Flashblock), or JavaScript. The only site that has been consistently slow is Delicious since those yahoos took over, and the only extension that has been slowing things to a halt is, ironically, NoScript.

Re:UI Lag

[spirit of reason]

I don't think that would work for people that play music in the background (via websites, that is).

Re:UI Lag

[hack slash]

That's where an exception rule/list could work, a similar setup to how FF deals with cookie acception/rejection.

Re:UI Lag

[hairyfeet]

Interesting. Are you using Readyboost? Because one of the first things I did with my main PC (Win7 X64, AMD 925, 8Gb of RAM, ATI HD4650) was to add a 4Gb flash drive for Readyboost and thanks to Superfetch it just flies. Now the Sempy is running XP and only has an SiS GPU and 1.5Gb of RAM, but I haven't found it lagging except in HD flash, which is natural since there is no GPU support. That is why I just use Downloadhelper when my quad is busy transcoding and I'm surfing on the Sempy.

Perhaps it is the way you have /. set up? I have images turned off in settings (no point in seeing Billy Borg or the Apple logo) and have comments set to -1. Who knows, maybe FF just likes AMD CPUs. It wouldn't be the first time I've seen an app favor one CPU over another.

Re:UI Lag

[CBravo]

Instead: IE just waits for all of the html to arrive before showing anything. In single-page-manuals that really nags me.

Re:UI Lag

[knarf]

Funny, that. I have tried Chromium on my state of the art IBM ThinkPad T23 and found it to suffer from UI lag and stutter way more than Seamonkey does. I keep on trying the most recent versions (now at 6.0.444.0) but the problem remains: open a tab, from within that tab open some links in background tabs and watch the whole thing stutter and halt regularly until the last background tab has finished its business - whatever that business may be as Chromium defers rendering background tabs until those tabs are displayed. Seamonkey (and its sibling Firefox) also have some of these problems but they are not as pronounced and aggravating as they are in Chromium.

Maybe Chromium works better om SMP machines (think multicore or multiprocessor) but as I don't have one of those since I retired my BP6 I can't say. On uniprocessor (or 'unicore' as that fits the bill better in these octacore processor days) configurations Seamonkey and its ilk are better behaved.

One annoying Gecko bug is that it sometimes manages to eat 100% cpu trying to resolve a name. That is a process which should not even show up on the processor statistics but somehow it does.

Re:UI Lag

[Hurricane78]

It’s interesting, how they are at fault, when you’re the only one with that problem.

I bet you set affinity once and forgot about it. And what sites do you open to get FF to 100% CPU anyway? I’ve only ever seen more than 50% CPU, when the Flash plugin caused trouble.

Re:UI Lag

[Hurricane78]

I must say that something is weird with the way Slashdot does JS. It’s only here, that having a page open in the background over time starts to rise in CPU and memory usage. As if a loop would constantly fork itself while not releasing its variables to free memory.

Re:UI Lag

[cerberusss]

He's not saying he has problems once the tabs are open. He's saying that when he starts Firefox, the browser opens the tabs from the previous start. And that takes a good 10-15 seconds, while the whole UI is unresponsive. That has been my observation as well.

Re:UI Lag

[Zuriel]

"Things work great so long as I don't ask it to do more than render basic HTML."

That's not really a glowing recommendation for a browser.

Re:UI Lag

[stewbacca]

What good is a responsive UI if it is a bad UI? Microsoft IE 7 (only using this as an example, because it's my current work browser), does everything it can to put stuff in non-sensical places. History and Favorites should not be found under the same icon, but if you must, the icon should make sense. A yellow star? Seriously? And when Microsoft is consistently last to the party with features (tabbed browsing, for example), what good is a responsive UI when it lacks features that all other browsers have?

I'm not meaning to rip on Microsoft, as I do like Win7 and the consolidated look and feel of the Office2007 suite (minus how Outlook Express still has the old interface until you open a new window to write or read a new mail, then you get the 2007 interface...weird, that). But IE is crap for about 4 versions in a row now, and I'm not even the type of person whoe cares which browser to use.

Re:UI Lag

[SimonTheSoundMan]

https://wiki.mozilla.org/Firefox/Goals/2010Q1/IO_Reduction

Seems they have been fixing your lag issues.

Re:UI Lag

[Anonymous Coward]

Trouble finding the right tab?

Re:UI Lag

[Anonymous Coward]

"The terrible user interface responsiveness of Firefox is what kept me on IE for the longest time (and I only moved because of addons, not because Firefox itself is any better)."

Well, that's about all one needs to know to gauge your expertise...

Re:UI Lag

[alder]

open a Slashdot story with ~1000 comments and watch as the browser just stops dead in the water for 5-15 seconds while it renders the page

I'd try to disable /. scripts (if you have NoScript). And maybe FSDN too... AFACT, and it was my experience, that the lag is not a page rendering time, but a script on a page trying to connect to a slow server and that, unfortunately, in FF blocks page rendering.

This is not a win-win solution :-) Some page functionality will be lost. Arguably not a very important part of it :-)

Re:UI Lag

[DigiShaman]

No Readyboost. Just two HDDs mirrored using the Intel storage manager. What I find both interesting (and frustrating) is that all instances of Firefox will freeze (not just tabs) until all Slashdot comments are rendered on a page. Depending on length, it could be anywhere from 2 to 5 seconds.

I haven't customized Slashdot much at all. Perhaps I should from the defaults if it annoys me enough.

Re:UI Lag

[dropadrop]

This is exactly the case. And what if you actually have to have one tab execute the shit that the grandparent is blocking (to make things work)? You end up with an unresponsive browser... I'm actually often in this situation, one tab processing something big and Firefox grinding to a halt until it's finished, and no plugins will help as I need to get it processed...

Re:UI Lag

[Anonymous Coward]

obscure WebKit/Gecko browsers usually don't have needed plugins like AdBlock

You serious?
Learn how to setup proxy filters, extensions are the awful-tier of media blockers. Proxies can completely remove any of the code or HTML you specify. (usually in simple RegEx commands, such as in Privoxy as a popular one)
Even HOSTS / IP Table blocks are better than extensions.
Custom DNS is just as good too, and allows for even better privacy.

Oh, damn it. CAPTCHA: bookworm.
Yeah, i deserved that.

Re:UI Lag

[drinkypoo]

Chrome's rendering engine is slow and sucks. Javascript is much faster, but that's it.

huh?. Chrome seems vastly faster in all the least-synthetic benchmarks.

Re:UI Lag

[Anonymous Coward]

Well, this is demonstrably wrong. I just opened this article (5,687 comments) in Chrome on Win7. All comments were retrieved. My machine: 2GB mem, 2GHz Athlon (IOW, ancient). Result: 200MB memory used by tab, tab was unresponsive for 15-25 seconds while it loaded. I could switch to another tab and interact with the page while the loading occurred in the background.

Re:UI Lag

[Malc]

Far worse than /.: we use Code Collaborator for our code reviews at work. Rendering those pages (e.g. two large C# files and their differences) can block the whole browser for several minutes. It's utter shit by FF.

Re:UI Lag

[Anonymous Coward]

Are you on Windows or Linux?

nmb3000: "...is what kept me on IE for the longest time"

Clearly a linux user.

Re:UI Lag

[bat21]

I'm terrible with opening too many tabs. I have a habit of opening 50 or 60 and then exiting the browser (but saving my session). The next time I'll open 50 or 60 more. I remember one time closing the browser and intentionally not saving my existing tabs, Firefox warned me that i would be losing 450 some tabs if I did so. I never had any performance issues unless I have a flash window that's acting up. Chrome on the other hand starts eating up my memory with only 50 or 60 tabs.

Re:UI Lag

[bat21]

That was the reason I quit using IE6. Firefox didn't have that terrible UI responsiveness issue on my machine, while IE certainly did. Well, I guess that and the tabs were the reason.

Re:UI Lag

[hairyfeet]

Are you talking about onboard RAID? Because that right there could be your problem. I would also look at chipset drivers. Working PC repair for nearly 15 years I have found that rarely does modern chipset drivers hard fail, they just give you little "hiccups", often with certain pieces of software: browsers (especially those that cache and fetch like FF does) along with CAD and Graphics software like Bryce or PS.

Try a previous driver for your chipset. I have often found when coming across these "little irritants" that going back one or two on the chipset driver will suddenly "make everything smooth" once again.

Re:UI Lag

[Dagger2]

If you're on 64-bit Windows, you can improve this a bit by using editbin (from the platform SDK) to set the large address aware flag in firefox.exe, increasing the limit to 4 GB. (If you're on 32-bit Windows, this will have no effect unless you also use the /3gb or /userva kernel flags.)

There are unofficial 64-bit builds, as well as ongoing work on official 64-bit Windows builds.

Re:UI Lag

[Anonymous Coward]

Does it? I didn't realise it did that without an extension like CookieSafe (I Just checked and couldn't find any option to do it on vanilla Firefox). Besides the behaviour you suggest is a little complicated for your average user, they probably wouldn't be aware why their streaming music or whatever keeps stopping after a few minute, not know how to make it work and then switch to a browser where it does work.

Re:UI Lag

[Anonymous Coward]

running IE through Wine.

Finally

[Anonymous Coward]

Whenever I open multiple youtube tabs, I cringe in fear that one of them kills the browser. HD video especially tends to do this. This really sucks if I've got a nice load of, um, art in other tabs and I lose them. Reloading might not even be feasible, due to the transient nature of today's websites.

Re:Finally

[Achromatic1978]

Why not just say "I surf 4chan for porn". Most of know exactly what you meant.

Can already kill Flash in 3.6.3

[kbahey]

I confused, since I am on Kubuntu 10.04 64-bit version, and use the Firefox version that comes with that release (3.6.3).

For the longest time, I am able to kill npviewer.bin without Firefox crashing. I just get a grey box when I do that where Flash used to be.

Flash already runs as a separate process for me.

Here are the processes:

me 4177 1746 0 12:43 ? 00:00:00 /bin/sh /usr/lib/firefox-3.6.3/firefox
me 4182 4177 0 12:43 ? 00:00:00 /bin/sh /usr/lib/firefox-3.6.3/run-mozilla.sh /usr/lib/firefox-3.6.3/firefox-bin
me 4186 4182 9 12:43 ? 01:03:08 /usr/lib/firefox-3.6.3/firefox-bin
me 4353 4186 2 12:45 ? 00:16:37 /usr/lib/nspluginwrapper/i386/linux/npviewer.bin --plugin /usr/lib/flashplugin-installer/libflashplayer.so --connection /org/wrapper/NSPlugins/libflashplayer.so/4186-1

So, what is happening here?

Re:Can already kill Flash in 3.6.3

[yuhong]

That is because you are using nspluginwrapper to wrap the 32-bit Flash plugin.

Re:Can already kill Flash in 3.6.3

[CRCulver]

You've also long been able to kill Gnash without taking the browser with it.

Re:Can already kill Flash in 3.6.3

[xSander]

As been said, you're using the 32-bit plugin with nspluginwrapper.

I've been using the 64-bit Flash for Linux for a while now. Up until a few weeks ago Flash would always crash on me after playing a video and closing the tab or going to another page, and I always had to kill the whole browser.

Funny

[Anonymous Coward]

My favourite part about this is that discusses Flash's severance from Firefox for all examples. Flash is taking such a beating lately. I'd probably still use Firefox over Chromium if it started as fast and didn't time out on pages like Google for no reason.

Haven't upgraded for a minor nitpick

[MortimerV]

This sounds like a nice feature. My flash and other processes don't crash on me often, but when they do they can be frustrating. There was one thing that kept me on firefox 3.5 rather than 3.6, though.

It's really silly, but.. when I use the autoscroll (middle click) and am slowly scrolling down through a page, I like to use the mousewheel to scroll faster occasionally, or back up a little bit, while the scroll is still moving. In 3.6 I found that moving the mousewheel canceled the autoscroll. This is also a problem because my mousewheel's a bit sensitive, so sometimes just brushing it would cancel autoscroll.

Anybody know if there's a way to change that behavior so mousewheel doesn't cancel autoscroll, or if it's been reverted in a later 3.6 release?

Opera!

[uid8472]

Has no-one else yet commented to point out that Opera has run plugins in a separate process for years now? Then I guess I have to.

Not to minimize the accomplishments of the Firefox developers, I mean, and getting this feature to the Firefox userbase is valuable in and of itself, and so on. But there is precedent.

Re:Opera!

[clintre]

Opera people always crack me up. Don't get me wrong Opera is a fine browser and may be better in this an that. However it it was really all that, it would have a much larger fan base. I just do not like it and judging by all the political moves they have to keep making just to survive it seems most others do not as well.

Just because they have had something for a while now, does not mean that Firefox, which is a far more popular browser, getting it is not a big deal. Opera has good plugins, but they have never been as good or easy to use. I think Chrome has already passed them up on quality plugins.

I do empathize though. If Opera would have made better decisions on distribution of the browser freely early on who knows where they could be?

Re:Opera!

[Nimey]

Opera is a poor imitation of lynx.

Re:Opera!

[luckymutt]

However it it was really all that, it would have a much larger fan base.

Popularity != better. Since IE has the largest fan base, you're saying that IE is the browser that is "all that?"

Just because they have had something for a while now, does not mean that Firefox, which is a far more popular browser, getting it is not a big deal.

Sure it's a big deal. Although it would have been a bigger deal if they were the first on the block to have gotten it.

Opera people always crack me up.

FF fanbois always crack me up. Do you people ever get tired of the pissing contest? Ever? And by the way, I am typing this in Konqueror. Suits my needs well enough.

Re:Opera!

[xigxag]

Popularity != better. Since IE has the largest fan base, you're saying that IE is the browser that is "all that?"

All other things being equal, the better software should be more popular. Why wouldn't that be the case?
Arguably, IE's market share is no exception to that principle...IE has traditionally been "better" for the average person simply because it comes pre-loaded on the OS instead of them having to try to find a legitimate download site. And it seems to me to be quite difficult for most people to distinguish malware from legitimate freeware/shareware. [Side note, I don't actually agree that IE has the largest "fan base." ]

But Opera vs. Firefox or Chrome, where's the disadvantage? Why can't it gain traction? Instead of playing verbal sparring games and gotchas, consider pondering that issue.

Re:Opera!

[Ndymium]

I would like to comment that it, in fact, doesn't. I've run Opera on OS X and Windows for a few years now and have seen no indication of that. In fact, I can see only one Opera process in Activity Monitor right now, with 15 threads - even if I open up a Youtube video. When Flash crashes, so does the whole browser (which used to happen all the time with the 10.5x betas). I've heard rumors on the My Opera forums that Opera on *nix might have this, but the OS X version certainly doesn't and I have no knowledge that the Windows version would either. Opera is a great browser, but this is something I've yet to see (and am eager to).

Re:Opera!

[Jugalator]

Popularity != better. Since IE has the largest fan base, you're saying that IE is the browser that is "all that?"

Neither Opera, nor Firefox or Chrome, are shipped with any Windows version.

Re:Opera!

[jo_ham]

Hold on, so when Apple adds multitasking to the iPhone everyone says "no big deal! Android has had that since launch!", but when Firefox adds separate process plugins which Opera and Safari have already done (Safari has been doing it since v4) suddenly it's "a big deal" because it's Firefox?

Don't get me wrong, it's good to see this (when Apple did it to Safari, the number of browser crashes I have seen has dropped *enormously* since flash only falls over by itself now). The GP was just pointing out that, but that Firefox isn't the be all and end all and that this approach is not new. You know, in exactly the same way people point out things like that in an Apple thread :)

Re:Opera!

[broeman]

Exactly, and furthermore Mozilla and Opera have had the same struggle for a fan base in the same period. Mozilla not only did good marketing, but they also managed to create a browser, that people really wanted to use (Opera just is to weird for some people, like me, although I liked Opera Mini on my old Symbian phone).

Re:Opera!

[Lisandro]

However it it was really all that, it would have a much larger fan base.

Opera IS a great browser, but your argument holds no water. It's a great product with a large fanbase - check the Opera forums if you don't beleive me. Firefox is great and a way much better browser than IE ever was, and check the usage statistics for both. Sometimes popularity has little to do with value.

Re:Opera!

[Anonymous Coward]

FF fanbois always crack me up. Do you people ever get tired of the pissing contest? Ever? And by the way, I am typing this in Konqueror. Suits my needs well enough.

Opera fanbois always crack me up. Do you people ever get tired of the pissing contest? Ever? And by the way, I am typing this in Lynx. Suits my needs well enough.

Re:Opera!

[Anonymous Coward]

+1

BTW, konqueror has separate process for plugins since ages. when flash crashes, just reload the page.

I pray the day were FF, flash and all those bloated software will disapear and be replaced by clean HTML5 rendering engines.
BTW, I4m happy that this bloated flash technology do not reach my iPhone. Adobe wan't to target iPhone but is unable (since 2 years) to port flash to x86_64 architecture. Hey adobe, start cleaing up your code and show portability by releasing 64bits flash and then we'll start to speak!)

As for FF, it's only advantage is it's ability to render badly written sites with average quality. Nothing more.

Re:Opera!

[Hurricane78]

WTF? Can someone fix this troll-moderation?
It seems someone just couldn’t stand the facts.

Am I the only one who noticed that the moderations here got worse and worse?
As if people would replace basic reasoning and logic with emotional rage and religious nutjobbery.

Protip: I’m using Firefox as my main browser. I recommend Opera to non-web-developers. And after having gone to 5 years of torture of having to develop for IE5 and 6, I stopped talking to people who prefer IE. If you had done it, you’d understand, and wonder why I’m still so nice. ;)

Re:Opera!

[ImprovOmega]

I downloaded Mozilla milestones between NN 4.5 and FF1.0. Milestone 13 added tabbed browsing and was a most welcome change from the very start. I haven't used an non-tabbed browser since 2000.

Re:Opera!

[Hurricane78]

Now let’s see who’s done first, troll moderator! You with your hand full of mod points? Or me with my excellent karma?
You know what? Fuck you. Fuck you majorly. Right where the sun don’t shine. Big old dinosaur dick, comes from the past, and takes you, right up the ass.

Go on. Mod me down. And use up all your mod points. I can do this all year long. Then at least you can’t damage others anymore. Go on. After all it’s the only thing you can do anyway. You wouldn’t be able to write a proper comment, if you life would depend on it. ^^

Re:Opera!

[Anonymous Coward]

Install base != Fan base

IE has the largest install base because it comes by default on Windows, that doesn't mean that everyone that uses it likes it, they might just not know about alternatives, not be able to change their browser (at work), or just not care.

Much smoother Flash video!

[steveha]

At work I have a Windows PC, and I was always frustrated by the very poor performance of Flash video. The video would freeze, then unfreeze over a second later with the video frames in between just dropped. (When you are watching a 5 second film this problem makes the movie almost unwatchable!) And it's a quad-core AMD Phenom II system. It should be fast.

So now, I'm trying out 3.6.4 and the difference is stunning. Now the Flash video playback is perfectly smooth.

I still want WebM in HTML5 instead of Flash, but what the heck, this is working now and I'm happy about it.

Sample size of one, YMMV, etc. But I'm happy about it.

steveha

Just as Adobe drops 64 bit Linux

[tsotha]

Just recently Adobe announced they would drop support for 64 bit Linux. What is wrong with these people? Is it really so difficult to put out a 64 bit version of software you already have running? Oh, but they promise they'll get it working someday. Thanks a lot, guys. It's a shame 64 bit computers are so damn new I have to use a wrapper to use your buggy, bloated, insecure, crap software.

Re:Just as Adobe drops 64 bit Linux

[jo_ham]

This is once issue where OS X users and 64 bit Linux users can huddle together for warmth in the cold and the snow. At least we sort of have a working browser plugin, if you have a powerful enough machine to brute force a simple task like an SD video, or a page of navigation links etc.

Re:Just as Adobe drops 64 bit Linux

[paulbiz]

And if you read the Adobe page linked to by that article, it says: "We are fully committed to bringing native 64-bit Flash Player for the desktop by providing native support for Windows, Macintosh, and Linux 64-bit platforms in an upcoming major release of Flash Player."

Re:Just as Adobe drops 64 bit Linux

[tsotha]

Like I said, they promise they'll get it working someday.

You realize they've been saying that since 2008, right?

Correction: Bugfix will be in 3.6.6

[behindthewall]

According to the discoverer and the issue; he mixed up two different fixes, initially:

http://lcamtuf.blogspot.com/2010/06/yeah-about-that-address-bar-thing.html

https://bugzilla.mozilla.org/show_bug.cgi?id=556957#c46

So...

[sootman]

... if Firefox crashes will all the plugins keep running?

Re:So...

[thoughtsatthemoment]

No. If I kill firefox.exe in the Task Manager the plugin process disappears too.

Re:So...

[sharkey]

If? Not when?

Re:So...

[oodaloop]

Ummmmm....whooosh?

Re:So...

[Anonymous Coward]

that .exe, what does that do?

Re:So...

[Anonymous Coward]

Informative? Anyone who knows how process management on modern OSes works would expect that to be the outcome (admittedly it is not a given, since the parent process could spawn the child processes in a way that doesn't result in this, but that is atypical behaviour).

What the fuck is happening to Slashdot where this crap get modded informative instead of redundant? Or better yet why isn't there a "Whoosh" mod?

Re:So...

[thoughtsatthemoment]

If it is not a given, then it is informative for those who really don't know. Even if you are a genius you would still try to make sure of it instead of just make assumptions.

"redundant" make sense here only if someone already posted that, but that's not the case. someone here already commented with a "whoosh" so your post is actually redundant.

single process for all flash

[thoughtsatthemoment]

It looks like there is a single process plugin-container.exe to run all flash files. Killing this exe will stop playing all the flash files. This means while you are enjoying a show on hulu.com, a rogue flash ad could still spoil the fun.

Re:single process for all flash

[Luminous Coward]

This means while you are enjoying a show on hulu.com, a rogue flash ad could still spoil the fun

You might try Flashblock

Flashblock is an extension for the Mozilla, Firefox, and Netscape browsers that takes a pessimistic approach to dealing with Macromedia Flash content on a webpage and blocks ALL Flash content from loading. It then leaves placeholders on the webpage that allow you to click to download and then view the Flash content.

FlashMute

[klui]

If you use FlashMute under Windows, edit HKCU\Software\InDev\FlashMute\filenames to include plugin-container.exe. I also found out the FlashMute volume slider control works under Flash 10.1.

Re:FlashMute

[alejobd]

Thanks. Worked like a charm

Firefox futures

[DragonHawk]

I'll take this opportunity to post some non-inflammatory info on planned Firefox development.

Firefox 4.0, which may go into beta as early as next month, is supposed to do a lot in this direction. Overhauled JavaScript engine, overhauled HTML rendering, etc.

http://wiki.mozilla.org/Firefox/4/Beta

http://developer.mozilla.org/en/Firefox_4_for_developers

I thought I had heard that 4.0 was supposed to deliver one-process-per-page functionality, but I'm having trouble finding recent status info. (One drawback to high-speed FOSS development is it's hard to keep track of things like that.) But anyway, the project is named "Electrolysis" ("E10S" in Firefox-developer-speak).

http://wiki.mozilla.org/Electrolysis

http://wiki.mozilla.org/Talk:Firefox/Roadmap

Re:Firefox futures

[surveyork]

http://www.product-reviews.net/2010/05/11/mozilla-firefox-4-beta-video-presentation/ Towards the end of the presentation, if my ears don't deceive me Mr. Beltzner says that there won't be electrolysis until 2011. It seems that they are having many problems with that. You can still try it in the nightlies or in experimental builds, I think.

Re:Firefox futures

[Anonymous Coward]

Please hand in your nerd ID card. No real nerd has used a graphical forward/back button in a browser since 1998. The only reason they still exist is for people over the age of 70 and under the age of three.

Re:Firefox futures

[Anonymous Coward]

Sometimes we have to use PCs that don't have a mouse with back and forward buttons.

Re:Firefox futures

[Anonymous Coward]

Don't forget the new HTML5 parser that is already working in the betas. Not only will this be the first fully HTML5 compliant parser, it will also be faster, run in a separate thread off the main thread, and make it possible to use SVG and MathML inline in HTML documents.

http://hacks.mozilla.org/2010/05/firefox-4-the-html5-parser-inline-svg-speed-and-more/

Re:Firefox futures

[Anzya]

Sorry you still fail. Real nerds don't need the keyboard to do things like that. Learn to use backspace or alt+left arrow for back and alt+right arrow for forward.
Everytime you lift your hand from the keyboard to the mouse you get -1 nerd point ;)

Re:Firefox futures

[bronney]

I agree. I only use alt cursor to go back and forth. My mouse4 and mouse5 also works that way. The only time I needed those graphical buttons is when I need to go 3 pages back at once.

Re:Firefox futures

[Draykwing]

Backspace or alt+left_arrow for back, alt+right_arrow for forward. HTH, HAND. Firefox for Linux disables backspace acting as back in the default build, but you can re-enable it by setting browser.backspace_action to 0 in about:config. The alt+arrow_key ones work in every browser I've ever used. (IE on Windows, Opera on Windows and Linux, Chrome on Windows and Linux, Firefox on Windows and Linux, Konqueror on Linux, Arora on Linux)

Re:Firefox futures

[fuzzix]

Not only will this be the first fully HTML5 compliant parser

Really, fully compliant with an incomplete and moving spec? That IS clever coding.

Re:Firefox futures

[bami]

My alt key is broken, you insensitive clod!

Also, I like right-clicking on the back button so that I can skip those stupid pages who decide that it needs to autoreload as the entire frame if something like google image search iframes a page.

Re:Firefox futures

[Anonymous Coward]

Are you seriously suggesting that I use a graphical web browser with a fucking keyboard??

I'm sorry, but YOU fail. Keyboards are for command lines, code and writing email. If you use it for anything else then you need to catch up with the past 2 decades.

Re:Firefox futures

[Anonymous Coward]

Learn to use backspace or alt+left arrow for back and alt+right arrow for forward.

And you need to learn efficiency. I trained a howler monkey to surf the web for me. He downloads all my torrents, updates my facebook page, twitters what I eat, and even sends me audible alerts on each successful first post he makes on slashdot. Why? So I can spend more time debugging a submissive AI engine while overcoming the design challenges of stretching exoskeleton over weather balloons for my fembot. You, sir, fail as a nerd. Not even worthy of carrying my sock stuffed jock strap.

Re:Firefox futures

[icebraining]

In Vimperator, you can do the same as in Vim: use numbers to execute actions more than one time.

For example, alt+o goes back, but 3 and then alt+o goes back three pages.

Re:Firefox futures

[Anonymous Coward]

That was awesome.

Re:Firefox futures

[stony3k]

Better still, use some mouse gestures

Re:Firefox futures

[jesser]

I think the markup-parsing part of the HTML5 spec is reasonably stable. Firefox is the first browser to implement the algorithm, so Firefox may have useful feedback for the spec.

Re:Let's All Kill The Opera Trolls

[MobileTatsu-NJG]

Just think how much better the world would be if that stinking pile of fail that is Opera and its 2-3 idiots who use the piece of shit browser didn't exist.

Well, to start with, you'd be reading about the world with a more primitive browser.

Single process for each plugin

[DragonHawk]

"It looks like there is a single process plugin-container.exe to run all flash files. Killing this exe will stop playing all the flash files."

FWIW, Google Chrome works the same way.

I'm not sure, but I suspect this *may* be due to design of the whole plugin concept. I would guess that the plugin concept assumes a single monolithic process for everything. There would be no need for an IPC facility. So I would guess Flash doesn't expect to find different windows running in a different process space. I know I've seen Flash objects communicate between each other; I presume that's done inside the plugin. If I'm right with my guess, using a different processes for each plugin object instance would be a giant compatibility problem.

Re:Single process for each plugin

[BZ]

You're exactly right. Flash assumes that all running instances of it share a single address space and uses various internal communication channels to have the instances talk to each other. The Chrome folks actually tried a process per plugin instance, and it broke too much stuff out there.

Re:Single process for each plugin

[renoX]

>to design of the whole plugin concept.

*Sigh*, and to think that the "plugins" were described as a big improvement: want to have 'flexible' software? Use plugins.
But they don't even have 'fault isolation'(*) right!!

WTF?

* and resources management and security.

Re:Single process for each plugin

[Malc]

I don't get it: what happens if I run something in IE and FF at the same time? Then there are two instances of Flash in separate address spaces. Killing IE doesn't bring down Flash in FF. What's going on in FF that doesn't allow it to run in multiple instances?

Re:Single process for each plugin

[BZ]

> Then there are two instances of Flash in separate address spaces.

Yes, but you don't expect the pages you have loaded in IE to affect the ones you have loaded in Firefox. If you have a page loaded in Firefox and you click a link that opens a new window from that page, the new window can affect the page it was opened from. In particular, they can share Flash plugin state.

Re:Single process for each plugin

[Malc]

That sounds like a problem with FF, not Flash.

Doesn't IE or Chrome make a decision about new pages as to whether they get a new process or share an existing one? It would seem like that logic would apply to plugins too. There's little reason for unrelated web pages to share a plugin, especially if they're from different domains.

Re:Single process for each plugin

[BZ]

> That sounds like a problem with FF, not Flash.

Why? A Flash instance on one page wants to change the state of a Flash instance on another page. The mechanism Flash uses for this only works if both instances are in the same process. Where does a FF problem come in?

> It would seem like that logic would apply to plugins too.

The Chrome folks tried it. Sites broke.

> There's little reason for unrelated web pages to share a plugin

Or a cookie store for subresources? Or a cache? Unfortunately, none of those statements happen to be false in practice... Witness all the sites that let you log in with your Google account, say; Flash has some similar things that require state to be maintained across all plug-in instances and maintains such state in its address space.

Now Flash could be changed to maintain the state in a broker process, but that would need to be a change in Flash.

Re:Single process for each plugin

[Malc]

Most of your points are irrelevant if I happen to be using two browsers. Things aren't broken if I do that today, so what's special about working with one browser? Adobe are hardly kings of outstanding implementations... they products should be sand boxed. Especially if it reduces the changes of XSS, etc.

Re:Single process for each plugin

[BZ]

> Things aren't broken if I do that today

The point is, they are. All sorts of stuff on the web doesn't work right if you try to use two browsers the way you describe. Now it may be that none of the sites you happen to use depend on the things that break, but you're not the only person out there.

TreeStyleTab

[DragonHawk]

With 128-150 tabs open

No offense, but I think you're doing it wrong.

I routinely open that many tabs. But then, I work in a dynamic environment where I'm often being asked to do a dozen things at once, including several open-ended research projects, plus a handful of web-based apps, plus casual browsing, reading news, etc. And Slashdot, of course.

I'll put in a plug for my favorite extension here: TreeStyleTab. Rather than limiting tabs to a linear strip, this gives it a 2D structure. When I surf, inevitably one thing leads to another thing, which leads to a site which leads to six more things. So I middle-click almost every link, and it all gets organized into a hierarchical history. It helps me organize my thinking when I'm researching something, especially when I don't know exactly what I'm looking for.

TreeStyleTab demo shot

It's got a million options. You can configure it for all sorts of things. You can even have it put a 2D tab strip across the top of the window, if you like that.

The lack of a working TreeStyleTab clone on Chrome meant I went back to Firefox. Everything else was great, but I can no longer do serious web browsing without TST, and so that killed Chrome for me. Yes, it's that important.

TreeStyleTab: Don't leave your home page without it.

Re:TreeStyleTab

[nmb3000]

I'll put in a plug for my favorite extension here: TreeStyleTab.

Interesting, I might have to give that a try.

When I surf, inevitably one thing leads to another thing, which leads to a site which leads to six more things. So I middle-click almost every link, and it all gets organized into a hierarchical history.

I usually break into multiple Firefox windows when I start having multiple disparate browsing sessions take place. I'll have a window for personal stuff (mail, Slashdot), a window for all the documentation I've got open, and maybe a third window for miscellaneous things.

The other thing I've really liked using is Session Manager because it makes saving these sessions really easy. I have around a dozen saved sessions related to various research projects I've worked on. Need to remember all the pages I referenced when looking at options for authenticating Linux off Active Directory? Just load the session of 20 tabs into a new window. It also makes restoring crashed sessions and whatnot a lot more flexible.

Nope, sorry

[yuhong]

"And reader Trailrunner7 supplies another compelling reason to download 3.6.4: "Security researcher Michal Zalewski has identified a problem with the way Firefox handles links that are opened in a new browser window or tab, enabling attackers to inject arbitrary code into the new window or tab while still keeping a deceptive URL in the browser's address bar. The vulnerability, which Mozilla has fixed in version 3.6.4, has the effect of tricking users into thinking that they're visiting a legitimate site while instead sending arbitrary attacker-controlled code to their browsers."" Nope, sorry: https://bugzilla.mozilla.org/show_bug.cgi?id=556957#c46

Browser process models and multitasking

[DragonHawk]

For performance reasons, tabs don't and shouldn't run in separate processes.

I find that statement dubious. Please explain.

In my experience, the process-per-page (be they tab, window, or whatever) yields much better performance. I believe there are multiple reasons for this. For starters, the OS already has a perfectly good scheduler, and it makes sense to use that to handle multi-tasking. Indeed, OS people prolly know more about how to design a scheduler than browser people. By exposing the this to the OS, it also means the OS can do whatever tricks it has to make I/O, memory allocation, etc., more efficient on a per-page basis, rather than treating the whole browser as an opaque object.

Finally, lot of modern hardware has 2, 3, 4 or more processor cores. Firefox generally only uses one of them. A browser like Chrome can have each page render on its own processor core, which is a *huge* performance gain. Without that, any multitasking is going to be limited to slicing up a single core between multiple tasks. The system can still only do one thing at a time. By using multiple cores, the system actually gets multiple things done literally simultaneously. On good hardware, the performance difference is astounding.

"You know, the original motivation for the tabs feature was that each tab could be run in a separate thread whereas each window needs a separate process."

That's just plain wrong. Each window does not need a separate process. Each tab does not get a separate thread. In Firefox 3.6, multiple threads are used, but it's not a one-thread-per-tab thing. Most of the work is still done in a single monolithic thread.

The motivation for tabs in Firefox was to copy Opera. The motivation for tabs in Opera was as an alternative to one-page-per-window or MDI.

Re:Browser process models and multitasking

[Khyber]

"In my experience, the process-per-page (be they tab, window, or whatever) yields much better performance."

While reading Slashdot, it doesn't make one bit of difference. While one story tab loads, the rest of Firefox FREEZES while slashdot struggles to get rendered. I can't even scroll up or down.

Which makes me think it's not a browser problem any longer, but the coders of websites and the coders of plugins (Crash er I mean Flash) that are the issue.

Re:Browser process models and multitasking

[EvanED]

I believe there are multiple reasons for this. For starters, the OS already has a perfectly good scheduler, and it makes sense to use that to handle multi-tasking. Indeed, OS people prolly know more about how to design a scheduler than browser people. By exposing the this to the OS, it also means the OS can do whatever tricks it has to make I/O, memory allocation, etc., more efficient on a per-page basis,...

What makes you think that threads aren't exposed to the OS?

Okay, there are userland thread libraries, but they're not what people are talking about when they say that each tab runs in its own thread or something. 99.9% of the time now, "threads" are either OS threads or userland threads implemented in a library that provides a rather higher-level abstraction than just threads, such as Cilk or TBB.)

Of course, Firefox doesn't actually increase the number of OS threads when you open new tabs, so one of three things is true:
1. The number of threads is independent of the tabs you have open
2. FF is keeping a pool of threads (either itself or a library like I mentioned above) and doing userland scheduling among them
3. FF violates what I said above and actually is using userland threads in a pretty dumb way (unlikely)

Regardless of what Firefox is or isn't doing, "letting the OS know" is most definitely not an argument to not use threads.

Re:Browser process models and multitasking

[keeperofdakeys]

If firefox is using userland threads, then the library they use must only be able to use one OS thread. This is a disadvantage in a computing world where desktop cores seems to be increasing in number and decreasing in power. Using a separate thread for each tab might be more trivial than trying to make the userland library OS thread aware.

Re:Browser process models and multitasking

[Anonymous Coward]

For starters, the OS already has a perfectly good scheduler, and it makes sense to use that to handle multi-tasking.

Modern OSes also have perfectly fine schedulers for threads that can handle multithreading.

Privilege separation, anyone?

[FraGGod]

Ok, now that we're able to put flash code in a separate proc, my question is: can we cut it's privileges so another (monthly) "zero-day vulnerability" will finally become just a tale to scare little children?
Strangely enough, with all the concern about flash security, article seem to miss that point.

Re:Privilege separation, anyone?

[BZ]

You can, if you're willing to break enough sites... Flash commonly performs network access, raw graphics operations of various sort, file access, and a few other things like that which would have to be disallowed in a sandbox.

Re:Privilege separation, anyone?

[FraGGod]

Well, I can think of a "nobody" user with home in, say, /tmp/nobody, and flash is running with it's uid and cgroup'ed, so:

• flash can read any libs or binaries (for these raw graphic ops, I presume) from fs as needed.
• flash can't access sensitive data in /home/myuser.
• flash can't write to /home/myuser/.mozilla/firefox/**/some_binary_file (that might get injected into process, run as "myuser").
• it can write to it's own "home" and access network as it pleases, although it will die along with a browser tab (cgroup gets killed, and flash can't escape it via forks).

I don't know much about what files flash accesses on local fs, but it certainly doesn't need write access anywhere but $HOME on unixes (works fine w/o it as it is), and I doubt it ever accesses ~/.mozilla (or ~/.opera, ~/.chrome, whatever) directly - these are subject to a constant change and shouldn't be necessary for the plugin which has direct interface to a working browser (whatever one it is). What am I missing here?

Re:Privilege separation, anyone?

[BZ]

What flash needs for a lot of what it does is raw device access. In Linux terms, access to stuff in /dev (video, camera, audio, etc).

It's clearly possible to setuid Flash to a low-privileges user if you want it to not write to disk in general and don't mind breaking part of the functionality. The question is whether you're willing to break it. Browsers may not be in a position to do that (though you individually may be if you don't use certain Flash features yourself).

Re:Privilege separation, anyone?

[jcupitt65]

It's things like webcams, audio, low-level graphics and sockets, as I recall. On IE8 (the one with a sandbox) the flash plugin is split into two parts. They have an in-sandbox stub which communicates over a channel to an out-of-sandbox process which contains the actual flash player. This out-of-sandbox player runs with full rights.

So even if FF did have a full sandbox for plugins, Adobe would probably make a hole in it for Flash.

TreeStyleTab and Session Manager

[DragonHawk]

I usually break into multiple Firefox windows when I start having multiple disparate browsing sessions take place

I used to do that, before I got TST. And I still do it sometimes -- especially if I've got different things happening on different virtual desktops. But I find the ability to expand and collapse tab branches is more flexible and more useful than static windows.

I actually suspect the tree-style-tab concept would be a good idea for a general purpose window manager, i.e., for all windows on a system, not just the browser. Might be tricky to figure out a general solution, though.

The other thing I've really liked using is Session Manager

I use and like Session Manager as well. TreeStyleTab and Session Manager play nice together; the TST tab structure is properly restored by SM.

No 64-bit version on the Mozilla website

[innocent_white_lamb]

Once again there is no 64-bit Linux version of Firefox available on the official download site.
 
Sigh...

Re:No 64-bit version on the Mozilla website

[Anonymous Coward]

stfu

Re:No 64-bit version on the Mozilla website

[BZ]

This is at least in part because on the 3.6 branch the 64-bit version is not at feature parity with the 32-bit one (for example doesn't have the JS jit, so has much worse JS execution performance). So linking to it on equal terms really doesn't make sense.

For 4.0, 64-bit Linux builds are much higher quality (for example they actually have the automated correctness tests run on them). So there's a decent chance those builds might become tier-1 by the time 4.0 ships.

Re:No 64-bit version on the Mozilla website

[TeknoHog]

It's available on Gentoo right now, I'm compiling it on x86-64 and ppc32. So not only is the source out there, but people have already tested the build on several architectures.

Re:No 64-bit version on the Mozilla website

[innocent_white_lamb]

Firefox 3.6.4 just showed up as an update to RHEL 5 / Centos 5.

https://rhn.redhat.com/errata/RHSA-2010-0501.html

There is both a x86_64 version and a i386 version available on the update mirrors.

Up to now, the latest Firefox on Centos 5 has been 3.0.something, so this is a big update.

Crashing

[boybarian]

It kept crashing on the NY Times site, until I disabled FireTorrent 2.0.1. I think this add on has caused me some problems in the past as well.

Firefox does NOT do process-per-page

[DragonHawk]

"In my experience, the process-per-page (be they tab, window, or whatever) yields much better performance."

"While reading Slashdot, it doesn't make one bit of difference. While one story tab loads, the rest of Firefox FREEZES while slashdot struggles to get rendered. I can't even scroll up or down."

That's because Firefox uses a single thread for just about everything. If a page is slow to render because of complex HTML/CSS, or has bad JavaScript which eats up CPU time, that drags everything to a stand-still.

Browsers that use a separate process/thread per page, on other hand, will keep everything else running. That one page will be slow/non-responsive, but everything else keeps humming along nicely (as long as the hardware can keep up). Google Chrome works this way. Firefox does not (yet).

(Firefox does spawn multiple threads, but the bulk of the work appears to be done in one thread. I presume the others are support/helper threads.)

Re:Firefox does NOT do process-per-page

[RebelWebmaster]

From what I understand, moving HTML parsing off the main thread with the new HTML5 parser will make a big difference in that respect for Firefox 4.0.

Firefox does not use process/thread-per-page

[DragonHawk]

"What makes you think that threads aren't exposed to the OS?

Um. I don't think that at all.

"letting the OS know" is most definitely not an argument to not use threads."

I wasn't trying to argue that. Indeed, I think it would be great if Firefox used a different process/thread for each page. I'm arguing *for* that. :)

We may be confused over terminology. When I say "multi-tasking", I am including both heavyweight processes and lightweight processes (threads). Also, I was looking at this from a somewhat Linux-centric point of view. On Linux, heavyweight processes and lightweight processes are very similar. They both use the same structures and system calls; it's just a question of what context they share. So the thread/process distinction is less critical than it is on OSes where a "thread" and a "process" are very different entities.

FF violates what I said above and actually is using userland threads in a pretty dumb way (unlikely)

On Linux, at least, Firefox uses multiple OS threads, but the bulk of the work appears to be done in a single thread. If multiple pages are busy, you don't see other threads picking up the load; you just see that one thread doing more work. I presume the other threads are helper/utility things, such as network I/O or name lookup.

Re:Firefox does not use process/thread-per-page

[EvanED]

I wasn't trying to argue that. Indeed, I think it would be great if Firefox used a different process/thread for each page. I'm arguing *for* that. :)

My apologies; it sounded like what you were arguing was "processes are better than threads, because (1) the OS's scheduler is better than what you'd write and (2) the OS can schedule the tasks on multiple cores." Your parent was comparing threads vs processes (saying threads were better because processes are more heavyweight), then you open up with "In my experience, the process-per-page (be they tab, window, or whatever) yields much better performance", so I thought you were arguing for processes over threads. ("Heavyweight over lightweight processes" if you prefer.)

On Linux, at least, Firefox uses multiple OS threads, but the bulk of the work appears to be done in a single thread. If multiple pages are busy, you don't see other threads picking up the load; you just see that one thread doing more work. I presume the other threads are helper/utility things, such as network I/O or name lookup.

Right; that was sort of my tacit accusation.

Re:Firefox does not use process/thread-per-page

[KingMotley]

Perhaps, but deciding to go one process per thread (almost) forces the developer to write code with a higher separation since trying to coordinate things across process boundaries can often be more difficult that implementing it correctly. In the end, this makes the code depend less on synchronization and scales much better.

It doesn't HAVE to work that way. You CAN write highly efficient multi-threaded code that has few dependencies and/or code/data sharing, it's just doing so across process boundaries is so much more difficult that it is often easier to rethink the whole problem and implement it so that the dependency and/or code/data sharing isn't necessary.

Separate processes

[h0dg3s]

"all run in separate processes, so when Flash decides to crash, it won't take your browser out with it."

That's what they said about Chrome. That was a lie.

Re:Separate processes

[swilly]

I have seen Flash die under Chrome and all I had to do was reload the tab in Chrome to get it to restart. I keep seeing people claim that Chrome is unstable, but I have never had any problems with it. Flash used to die quite often, but since the 10.1 update it hasn't died on me once. I'm pretty happy with Chrome + Flash 10.1 on both Windows and Linux.

I wonder what you are doing if you are seeing Flash kill Chrome like that.

Re:Separate processes

[h0dg3s]

Multiple Youtube windows open.

Hmm...

[clone53421]

If Flash starts consuming all the CPU it can find, you can kill it without nuking your browser session.

Sold! I’ll take it.

Java was already sort of its own process. Making other plugins do this as well will be a very good step.

Comment removed

[account_deleted]

Comment removed based on user account deletion

And people say Flash is consumer unfriendly

[gig]

So all we have to do is send all Web users to night classes on process management so they can diagnose when Flash is consuming too many resources and identify and kill the relevant process. That way we can rescue Flash designers from having to learn HTML and Adobe from having to compete with anybody. Makes total sense. I mean, playing video ought to be complicated, right?

 

Re:And people say Flash is consumer unfriendly

[softegg]

Yeah, because all anybody does with Flash on the web is play video. Nobody writes games, applications, or animations using Flash at all. Nope. Just videos. Yeah, everything should be done with open standards like HTML 5. Because, as everybody knows, HTML and CSS run exactly the same in every browser.

Yeah, that will work great...

Laptop Sleep Cause Flash to hange in FF

[DoChEx]

I've been running FF 3.6.4 for a 2 week now, download a beta, now the release one and on my laptop when I have a couple of YouTube videos open, if I put my laptop into sleep mode and when I come back all the pauses videos have frozen, they loop 4 frames.

Don't know if this is Flash or FF as both updated around the same time! It's very annoying

Cool multitasking

[Snufu]

Now I can watch Youtube and post on Slashdot at the same t

awesome.

[smash]

This is going to make me give firefox another shot. I've been driven to chrome for per-process tabs, and Safari for the eye candy (visual preview of bookmarks/history) - and firefox has just been this browser with the UI that is prone to lock up when something shits itself for me. Sure there are plugins but i can live without them for 99% of browsing I do.

Splitting plugins into a seperate process will be a massive win for UI response I reckon, downloading the update now :)

how to apply this fix?

[Anonymous Coward]

that buy says fix is https://hg.mozilla.org/releases/mozilla-1.9.1/rev/8c17cad9e838
and its to browser/base/content/browser.js
but there is no browser.js in my tree
update: oh right, https://hg.mozilla.org/releases/mozilla-1.9.1/raw-rev/8c17cad9e838

Mozilla Firefox\chrome\browser.manifest
says its in
Mozilla Firefox\chrome\browser.jar
so all you gotta do is
unzip -d browser browser.jar ...
rezip browser.jar, replace browser.jar and you're done .... why doesn't firefox issue an update, its just javascript?

Re:how to apply this fix?

[Anonymous Coward]

It seems to work, that is, URL isn't set anymore, url location bar shows nothing.
If you click url location bar, then Esc, it shows about:blank
party time i want your skull

Not Mac

[ischorr]

A tag on the post implies it, but I thought it'd be worthwhile to mention specifically that this applies to Windows and Linux, but NOT Macintosh.

Personally, although the Flash plugin for Mac is dramatically less speedy than on Windows, I've never had stability issues with it. I've never once, in the last (...6 years?) had it crash and/or take the browser with it.

My only real problem with Firefox has been the bizarrely high CPU utilization and tremendous memory leaks, neither of which are caused by extensions or plugins. For years now if I use the browser for a few hours or more, opening 4-15 tabs, it'll start using 30-50% of CPU. Certain sites seem to trigger it more than others, but it's not consistent. The problem persists even if I close all tabs, and happens even with Flash and other plugins/extensions disabled. In fact, it's hovering around 20% CPU usage right now and 600MB RAM consumed, and the only thing I have open is this comment window. I've had this problem now on 5 different Macs and various profiles over the years. CPU problem doesn't happen on Windows (but the memory usage problem does). And it's driven me to Chrome for the last 4 months, where at least memory is released/CPU drops when I close a tab (kill a process).

Re:Not Mac

[Jesus_666]

Apparently Mac users get out-of-process Flash not before Firefox 4.0. So I just hope 3.7 in its extension-breaking glory will be released soon so I can tell YouTube to use WebM and avoid Adobe's horrible not-designed-for-video-playback video playback technology as much as possible.

Re:Not Mac

[BZ]

3.7 got renumbered. It will be released as 4.0.

Re:Not Mac

[Jesus_666]

Well, that explains a lot of things. Thanks for letting me know (although I should have become suspicious when I noticed that 3.7 and 4.0 are "both" slated for release in 2010).

Doesn't seem work for Adobe Acrobat

[ctusch]

As far as I can tell the out-of-process feature doesn't work for the Adobe Acrobat plugin.

Increasingly buggy

[dugeen]

More stuff is broken with each new version, I notice - I trialled the last release, and not only were half the config settings not preserved from the previous version, but the browser is still refusing to download executables because Internet Explorer is set not to do that. Keep your nose out of my IE settings and kindly attend to your own, they're different for a reason.

Fuck flash

[Anonymous Coward]

It's a steaming pile of shit, just like 99% of its content.

Fastest browser?

[Anonymous Coward]

I'm tired of these people talking about shaving a billionth of a second off of initial page loading and somehow expecting that to mean something to me.

I've been using Firefox since it was called The Phoenix browser, and the same problems are always there. The UI is slow, and the scrolling is jerky as fuck. This browser chokes on simple HTML pages that would have scrolled fine when I was browsing on my 486 with IE4. In every other browser I can smoothly scroll pages that have flash objects an scripts running all over the place, but FF chokes on plain HTML pages that are nothing but a few icons, and an HTML table structure. It's almost like FF is clearing out RAM and then reloading the whole page from scratch every time I flick the mouse wheel. It gets 10x worse if page zooming is active.

The improved download manager is also full of the same old problems. I have my screen set to dim every 10 minutes if I've been inactive. The PC doesn't sleep, or hibernate, or anything, just a dimmed screen. No other program or browser cares about this, but the FF download manager decides to cancel all my downloads and act like my PC is in hibernate mode.

The program doesn't honor the settings for what to do when certain file types are encountered. I don't know how many fucking times I've had to set it NOT to open PDF files, but it keeps doing it anyways. I finally just got rid of acrobat all together. Not sure how you can call this a secure browser when it opens it's legs up and accepts every invalid, malware injecting PDF file on the net.

It also doesn't follow the update settings. Automatic updates are nice, but not when I've selected the option to not even do the check. This is really bad on a limited user account, where FF will automatically disable itself until someone with admin privileges can come and enable the program to finish the update. (this may have been fixed recently, but it was a problem for years...)

Firefox needs to stop setting meaningless metrics to call itself the fastest browser. I can make a side-by-side performance video too. It will show IE, Chrome, Opera and anything else scroll through and interact with the most CPU hogging sites on the net, while FF is chugging trying to scroll a .txt file.

I only use now because I've gotten used to some of the extensions and behavior. But as a browser, it's not actually good at any real world usage scenarios anymore. It has some nice perks when it comes to configuring misc behavior, which is why I ditched Chrome after trying it. But it's the worst browser by far when it comes to actually interacting with the web.

Re:Fastest browser?

[Hal_Porter]

Firefox needs to stop setting meaningless metrics to call itself the fastest browser. I can make a side-by-side performance video too. It will show IE, Chrome, Opera and anything else scroll through and interact with the most CPU hogging sites on the net, while FF is chugging trying to scroll a .txt file.

I find Firefox sluggish too - particularly recent versions compared to recent versions of Opera. IE used to be fast but almost as bloated as Firefox.

Well at least we know such a bug exists

[Ilgaz]

I use Opera all the time and I can't thank them enough for supporting Symbian S60 with a decent browser.

The issue with Opera, Safari (it is way more than Webkit shell), IE is: Who knows how many of such issues exist on them? It is more frightening if some gray/black hat found a similar issue on them and put it on black market.

Recently Opera released a security update to 10.53 (10.54) and declined to tell the "major" "critical" issues they fixed. I can't blame them, it is how they work. On the other hand, with Firefox, we _know_ about the exploit, AV vendors too (they can add to heuristics), system admins can take measures against it.

Don't get me wrong, the emphasis to security on Opera browser is amazing, that is one of the reasons why it doesn't have "too powerful" extensions and Widgets will never have the same power as Extensions. It is just, nothing can beat Firefox on "openness" and community. Just because Webkit/Chrome etc. are open source, it doesn't make them same as Mozilla Firefox.

What if you block Flash?

[stewbacca]

So now when you slashdotters launch the new Firefox, you get two processes (Firefox, and the plugin-container), but you all have blocked Flash, so you are running a second process for no reason. That doesn't seem very efficient. (hint: I'm trying to lighten my tone, so this is an attempt at humor).

Re:What if you block Flash?

[paulbiz]

The plugin-container is only launched when a plugin is actually loaded.

Semantic confusion

[DragonHawk]

"My apologies ... I thought you were arguing for processes over threads."

Yah, I can see where you got that from, now. No need to apologize; my fault for being confusing.

"parent was comparing threads vs processes (saying threads were better because processes are more heavyweight)"

At this point, I'm not sure *what* BhaKi was saying in #32661562. My original take was "tabs don't and shouldn't run in separate processes" meant any attempt at multi-tasking browser stuff should be avoided. But I see what you mean; they might have been arguing in favor of threads over more heavyweight processes. The whole tab=thread/window=process assertion threw me.

SE Linux issues

[ogre7299]

I would love to install this on my RHEL5 office system, but our IT guy has SE Linux enabled on all the systems. The new firefox throws AVC denials so I guess I'm stuck with older versions.

  and we don't have root access.

As a Linux user

[Anonymous Coward]

As a Linux user I can sum up my choice of browser in the following way:

Opera: Excellent browser. Has the best set of features of any browser out-of-the-box, almost no rendering issues and it's fast. Unfortunately it can't be patched, updated and packaged as easily as other free software browsers. It's closed nature also makes it non-portable, limiting me to whatever platforms Opera Software decides to support. It was my browser of choice for a long time but when I started to migrate to pure64 Linux Opera's releases didn't keep up. Ruled out.

Chromium: Also excellent. Unfortunately Google's development model for the browser makes it painful to package and distribute. The bootstrap tarball is a whopping 700MB in size, and after the tarball has been downloaded you have to update it with svn. AFAIK there are no regular release tarballs and shipping a 700MB non-current tarball in the source tree with a strange build system and code that has to be updated before building is out of the question. It would be my browser of choice on Linux if it didn't complicate things so much. I think most Linux distributors agree with a number of these points, which is why we don't see more of them package Chrome(ium). Distributions like Slackware would never, ever carry source code that big (at best you get the pre-built binaries from Google.. again, this affects portability and from what I know it's heavily optimized for x86, probably won't even work on PPC/ARM). Ruled out.

Konqueror: Great browser for the most part, but uncomfortable to use. Has rendering issues (and "flickering" when it draws and loads webpages, forms are sometimes broken etc.) which makes it annoying, plugins don't always work (like flash). And the way bookmarks is implemented isn't as polished as one would hope. KHTML is a good engine but not as good as WebKit, and QT's internal WebKit engine apparently needs work (based on my experience with Rekonq which needs a LOT of polish). Ruled out.

Epiphany: Haven't used the new WebKit-based version because I don't use Gnome (and it's heavily tied to it). Probably what I would recommend and use myself if I didn't prefer KDE as my desktop. Ruled out.

Firefox: The browser I prefer. It isn't the fastest browser but it's fast enough. It's easy to build and the functionality it lacks can be added with extensions. I use it because it's well supported and just works. Fact is, while there are plenty of browsers that can compete with Firefox in terms of features and polish (even exceed it) those aren't the reasons I actually USE firefox. It's might be based on Gtk but isn't tied into Gnome so it's well suited for using on desktops other than Gnome (like Xfce and KDE).

I don't fit the profile of your average Linux user so my reasons for choosing a particular browser is different from the norm, but the fact is that Firefox is good anough and it fits the free software development and distribution platform very well making it easy to support.

Re:First update, the new dead-fox maker

[Runaway1956]

I notice that you post as AC. I wonder - is your post just sour grapes? I've been running Namoroka for a long time now - currently at 3.6.6prerelease I think it's pretty sweet. Yeah, because I constantly run near cutting edge betas, there are glitches, but so far, it has NOT BROKEN.

How do you figure that FF doesn't test their browser, before release? I'm testing it! I've been testing it! Not just me, but thousands of other people who like to live life in the fast lane.

Come back and post a real complaint - maybe someone can help you - or not.

Re:First update, the new dead-fox maker

[tippen]

I ran into this as well. Let the auto-update move me from 3.6.3 to 3.6.4 and FF can't connect to *any* webpage now. Have had to resort to using IE until I figure out what the problem with FF is. Jay

Mod parent up. Firefox memory leaks.

[Futurepower(R)]

"I'm really hoping that separating the tabs will help with memory leaks."

Agreed. Somehow Firefox not only leaks memory, but corrupts Windows XP SP3 when it begins to need virtual memory.

Chrome hosed for me today also

[billstewart]

I decided to open the week's worth of FARK stories into Chrome today rather than Firefox, that was a couple hundred tabs. The great thing about running on a dual-core machine is that if Firefox gets hosed, it can only burn up one of the CPUs with runaway Javascripts, and the rest of the machine keeps working fine, and I was hoping Chrome would be better behaved. Well, Chrome got hosed and started burning 50% of my CPU capacity with runaway Javascripts or something, and was unhappy when my wireless connection died underneath of it, and more unhappy when I took the machine in to work and plugged it in there behind a firewall. It doesn't help that I don't have Chrome configured with all the Ad-Block, No-Script, and Ghostery that's protecting my Firefox from overbloated websites, but it still shouldn't have done that.

My Firefox borked on update also

[billstewart]

I had to reinstall it to get it to work. (Didn't help that the reinstalled version had an incorrect proxy setting on it, so it wasn't loading anything from the net until I found that :-)

Look at your network proxy setttings

[billstewart]

I'm not sure which of the chain of things going wrong this morning was caused by the upgrade dying in the middle (:-), but at least one of the problems was that the proxy setting got set to something that didn't match my network at work (which normally does transparent proxying, so I don't tell FF to use a proxy server, and it was trying to load one automagically or something.)

Re:Look at your network proxy setttings

[tippen]

We don't use a proxy, so that shouldn't be it but I can't check. I re-installed 3.6.3 and everything is working fine again. 3.6.4 == broken

Re:Look at your network proxy setttings

[billstewart]

We don't use an explicitly-set-in-the-browser proxy either, which was one reason that whatever proxy setting didn't work.

Parent is not a troll

[Weedhopper]

WTF is going on with moderation these days?

The above is an opinion and not an inflammatory one at that. It's somewhat rational, even if you disagree with it.

If I'd have had any moderation points after the AJAX debacle of the past few days, I'd correct this immediately.

Flash video + OOP == 100% CPU

[ThePhilips]

If Flash starts consuming all the CPU it can find

With the OOP feature it started doing precisely that. Very same behavior as in the Chrome - and one of the reasons that I do not even consider seriously it.

Gotta find a way to disable the enhancement soon.

P.S. I'd rather have Mozilla include FlashBlock instead of that by default.

Tradeoff for stability vs. speed going "OOPS"

[Anonymous Coward]

"Firefox post. Firefox is the fastest browser around!" - by Shikaku (1129753) on Tuesday June 22, @11:13PM (#32661222)

Typically, that's been OPERA's "informal title" over the past decade or so...

However, this movement of FF browser addons into "out-of-process" (external to FF itself in other words) libs being marshalled is going to lead to less speed in FireFox!

(Though it will aid overall browser stability, this IS the "trade off" here - yes, you shouldn't be able to crash the calling parent process, in this case the browser, but, you incur more messaging overheads).

In fact, I'd say "ask the FF devs" about what I just stated... odds are, they WILL "second my motion" here.

APK

P.S.=> A "sad fact of programming life"? It's trade offs like these, & there's no real way around them with things structured as they are & you can't win for losing (you appease the stability crowd, & then you have the speed-freaks to contend with afterwards (who the latter are the group you SHOULD put after stability & security, which it seems the FF dev team has done & rightfully so))... apk

Add OPERA SECURITY to your "repetoire" cgomezr

[Anonymous Coward]

You might want to keep these statistics in mind regarding Opera's FANTASTIC security trackrecord, because as per usual, Opera shows NO KNOWN SECURITY VULNERABILITIES left unpatched... see stats from SECUNIA.COM below, for the "big 4" webbrowsers:

---

Vulnerability Report: Opera 10.x (06/24/2010)

http://secunia.com/advisories/product/26745/

Unpatched 0% (0 of 7 Secunia advisories)

---

Vulnerability Report: Mozilla Firefox 3.6.x (06/24/2010)

http://secunia.com/advisories/product/28698/

Unpatched 40% (2 of 5 Secunia advisories)

---

Vulnerability Report: Microsoft Internet Explorer 8.x (06/24/2010)

http://secunia.com/advisories/product/21625/

Unpatched 31% (4 of 13 Secunia advisories)

---

Vulnerability Report: Google Chrome 5.x (06/24/2010)

http://secunia.com/advisories/product/30134/

Unpatched 0% (0 of 1 Secunia advisories)

---

("Read 'em & weep...")

APK

P.S.=> Keep this set of URL's in mind for your next posting in regards to Opera cgomezr... it always gets the "firefox fanboys" into an "uproar" every time I post it (which only shows you're getting thru to them in their zealotry via documented facts is all), & they eventually end up doing their "patented name-tossing ad hominem attacks" directed my way in the end, rather than attacking the documented stats above... LOL!

In fact, it reminds of a film I have just watched called "The Book of Eli"!

(Specifically in the scene where Eli has just come under the highway bridge underpass, & when he cuts off the cannibalistic hijacker's (troll) hand, & the rest of the band of his "hijackers" go to try to take on Eli (the FAT one makes me laugh - "King Pork" with a Darth Vader gas mask & a chainsaw), & they fail, badly because they are outclassed & overmatched...))

So, when I post this? Well, I think of it THIS way, per a quote Eli himself used in the film:

"Yea, though I walk thru 'the valley of the shadow of /.', I shall fear no "troll" (evil): For thou art with me" (the 'thou' simply being verifiable facts & figures from a respected website that specializes in being a clearinghouse for known application security vulnerabilities)... apk

Isolating Flash objects

[DragonHawk]

Most of your points are irrelevant if I happen to be using two browsers. Things aren't broken if I do that today, so what's special about working with one browser?

The key point you're missing is that some websites open multiple pages, instantiating multiple Flash objects. This is commonly done with popup windows, for example. Flash also has mechanisms for allowing different servers/sites/pages to share data and communicate with each other. Those things all work together, as a group. Such things will always be opened in the same browser, because they're opened by Flash, not by you clicking the Internet Explorer icon. Since historically, all plugins ran in the same process, the design of Flash assumes all the objects instantiated across all those pages will be able to communication within the same process. If you isolated each page's Flash objects in its own process, then all of that would break.

There's a very good chance you use web systems like this now, but just don't realize the systems would stop working if every Flash object within your browser was isolated.

Re:Isolating Flash objects

[Malc]

I was under the impression that that these multi-process browser implementations keep such related pages in the same process space, for the same reason (breaks JavaScript/DHTML/etc). It seems like plugins would be covered by this.

Firefox goodwill being squandered

[cavebison]

I think Firefox, as a FOSS project, is suffering from the "someone else's boring tasks" syndrome.

There are three major issues which have been outstanding for *over a year* which I, as a user, would rather have had fixed than all the fancy and I'm sure fun work being done on speed, OOP plugins, tabs and "features". Let's not even mention the AwesomeBar.

1. Bug 505455. Mouse events firing on background tabs. A year old, not fixed.
2. Bug 461483. Including "www." when searching in the the url bar. Over 18 months old, not fixed.
3. Non-standard Windows file dialogs, making FF3 incompatible with utilities like Direct Folders.

These things are why I went back to FF2, which works properly in all three respects, though it renders slower. Big deal. I use Chrome as needed for more advanced sites or when speed is a big issue. It's usability I'm interested in and why, sadly, I simply cannot use FF3.

But I like Firefox, as a product. And not just for the plugins, which have given FF a lead which, in my option, is being squandered by messing with usability and ignoring important but boring-to-work-on bugs.

Re:Let's All Kill The Opera Trolls

[Anonymous Coward]

Imagine that, a death threat from a troll. Cmdr Taco should report this to the proper athorities as this crosses the line on free speech. As a matter of fact, you should be charged for a felony and locked away for quite some time.

Process-per-page != process-per-plugin-instance

[DragonHawk]

"I was under the impression that that these process-per-page browser implementations keep such related pages in the same process space, for the same reason (breaks JavaScript/DHTML/etc)."

Nope. The code for the browser is under the control of the browser team, so they can adapt/rewrite things as needed for a multi-process model. They have no control over plugin code, so they can't do that.

FWIW, the commonly-cited reason for Mozilla Firefox still using a single process for all pages is that it's a lot of work to adapt all the existing code for a multi-process model. They're not there yet.

(Process == heavyweight or lightweight (thread) in the above.)