Multi-process Firefox brings 400-700% improvement in responsiveness

https://techcrunch.com/2016/09/02/multi-process-firefox-brings-400-700-improvement-in-responsiveness/

by bpierre • 10 years ago

572 337 10 years ago

gnicholas• 10 years ago

Instructions for enabling this feature (via about:config) are provided in the author's prior article on the topic: https://techcrunch.com/2016/06/10/mozillas-multi-process-arc...

daenney• 10 years ago

Just so that people don't have to do digging in that article. In about:config, set browser.tabs.remote.autostart to true and check about:support for Multiprocess Windows (might need to restart).

For now though you'll have to disable any plugins for it to work which can be a bit of a handicap. But... almost there :).

mintplant6• 10 years ago

Please be aware that if you force-enable e10s while using incompatible addons, you'll end up with a very slow browsing experience. http://arewee10syet.com/ has a compatibility list for some of the most popular addons.

AstralStorm6• 10 years ago

Got to love the '401 failed fetching from AMO' add on.

diggan• 10 years ago

Seems like e10s gets disabled if there is a incompatible addon activated.

Not sure I would trust that list, the only two addons I run, is Tree Style Tab and Greasemonkey, both listed as compatible on the website, but going to about:support shows:

Multiprocess Windows 0/1 (Disabled by add-ons)

sandyarmstrong• 10 years ago

What about using incompatible addons triggers a slow browsing experience?

Zekio• 10 years ago

The speed is great as long as youtube doesn't autoplay a new video which freezes the whole browser for me until it is done loading.

akerro5• 10 years ago

Whole browser freezez for me on any amazon page. Just when it starts rendering it freezes everything inside Firefox window for a few seconds. It's so annoying that I'm considering dropping Firefox after over a decade for Opera or Brave browser.

nl3• 10 years ago

I wonder if this is the same issue as I'm seeing: https://news.ycombinator.com/item?id=12437130

Amazon works for me though. Hmmm.

malnourish2• 10 years ago

Have you tried disabling the functionality in about:config?

mintplant6• 10 years ago

What addons do you have installed, and what OS are you using?

firebird84• 10 years ago

For what it's worth, you can turn off autoplay in about:config.

malnourish2• 10 years ago

Is there a list of commonly used extensions that are (in)compatible?

wjoe2• 10 years ago

http://arewee10syet.com/ has a list of addons and their compatibility.

I tried turning on the about:config option and the URL bar broke, so obviously something I have is incompatible - I suspect it's Classic Theme Restorer, although that site does list it as compatible.

danblick• 10 years ago

Does anybody know more about how Mozilla manages experiments? I'd always assumed that all users downloaded an identical binary and got the same behavior. How do they assign 10% of their users to an experimental group?

Is it possible that two users will see different functionality when they are both using the same OS and release, say Firefox from the latest version of Ubuntu?

TheCoreh• 10 years ago

As with most native A/B tests, the binary probably ships with code for both functionalities. When the browser launches for the first time they generate a random number, and if it's below a certain threshold you get variant A, above you get variant B. The information gets persisted on disk so it remains unchanged across sessions.

theandrewbailey2• 10 years ago

It seems that there is some communication back to Mozilla, since they claim that they can increase or decrease the number of users in each group over a week; that seems unlikely to come in the form of a new binary.

kbrosnan• 10 years ago

Firefox generates a random number on first run. Users who are selected for the experiment are sent a signed extension from Mozilla that makes browser.tabs.remote.autostart preference return true if they are on the correct release of Firefox.

https://wiki.mozilla.org/Electrolysis/Experiments

sp3324• 10 years ago

Certain features are rolled out with Test Pilot https://wiki.mozilla.org/Test_Pilot but that doesn't seem to include the e10s testing.

Sylos6• 10 years ago

Yeah, Test Pilot is more for features where Mozilla wants to experiment and see, if people even like the feature in general, not so much for bug-testing.

With e10s, it was clear that it would be included in Firefox, so they just tested it by pushing it through the normal chain of Nightly->Developer Edition->Beta->Stable.

lxt• 10 years ago

We have several different mechanisms for running experiments. The e10s one specifically runs through what we rather poorly named a "system add-on".

e10s code is present in all builds. The system add-on contains logic that decides whether to turn it on or not. The system add-on can be updated daily (or more often), in all channels.

* I apologize for the naming, this is entirely my fault. I spent a lot of the last twelve months scheming/designing and developing experimental update mechanisms for Firefox.

nl3• 10 years ago

I switched to Firefox from Chrome about 12 months ago. It isn't as good as Chrome, but I was trying to reduce my Google dependencies.

It has been mostly fine (except for an annoying OS-X multi-screen bug where it screws up the sizing).

I was really looking forward to this feature to help close the gap on Chrome performance.

Until August (I think Firefox 48.x), when it became unusable on any site with... something. I'm not entirely sure what triggers it- I don't think it is just video alone. Something make the entire browser lock up entirely for minutes, and sometimes it even runs out of memory and I have to kill it via the OS.

No add-ins (except for Firebug).

Frustratingly, I can't replicate it well enough to be a useful bug report.

I'm this close to switching back. Muscle memory and shortcuts to Firefox is the only thing stopping me.

So.. this will be great, but please make it a workable browser.

wjoe2• 10 years ago

I've found Firefox performance to be absolutely awful at work on OS X. I have no issues at all with it at home on Linux, and didn't have any issues in my previous job on OS X either. I'm not sure if it's OS X specific, a change in recent versions, this laptop being underpowered, or the nature of the work I do (a lot of video stuff), but anything JS heavy severely slows down or locks up the browser. And if anything, it's gotten worse over the last few months.

I have to use Chrome for anything video related or JS heavy (monitoring/graphing stuff for the most part), but I'd hate to switch to it for everything. There are too many addons that I rely on in Firefox, on top of my general dislike for Chrome's minimalist UI and standard Google privacy concerns.

I've tried turning on the config option for electrolysis, but it breaks some addons, so that's still a no go for me for now.

stcredzero2• 10 years ago

Firebug?

mintplant6• 10 years ago

Do you get the same problem if you temporarily disable your addons? It could be a bug in one of them rather than Firefox.

sfink• 10 years ago

Any chance you could file bugs, particularly on the JS-related performance issues? I'm looking into some similar things right now, but seemingly JS-centric things seem to have a tendency to actually be from something else. It'd be nice to have something I could solidly blame on JS. If you file a bug with https://bugzilla.mozilla.org/enter_bug.cgi?product=Core&comp... and make it block bug 1299643, I'll see it.

sp3324• 10 years ago

It's really hit-or-miss for me too. On my work laptop for example, it would freeze up for multiple seconds with e10s enabled but was fine with it disabled. Then one day everything worked fine. On my home computer FF is just getting slower and slower. I'm not looking for troubleshooting tips here, just mentioning that the experience is very uneven.

merb• 10 years ago

I've thought exactly the same.

mrweasel2• 10 years ago

For a while now I've suspected that memory management on OS X is actually terrible. I have no really proof, because I just run everything memory intensive on Linux. At work I have a few Python script that eats a few gigabytes of memory, it reads in data, and basically just loops over it, selects the stuff it needs an generate a few XML file (10 to 20MB worth of XML).

All this takes maybe five minutes on Linux, on Mac OS X it will freeze the OS for long periods of time and take 40 minutes to run. In terms of hardware, the Linux box has less RAM, CPU and slower discs.

Regarding the Javascript stuff: Part of it may be that Chrome is the new Internet Explorer. People develop on and test for Chrome (because everyone uses Chrome, right?). Chrome does have the better Javascript engine, still, I believe that much of the slow Javascript on Firefox (an other browsers) could be avoided if people bothered testing a little more.

enraged_camel• 10 years ago

>I've found Firefox performance to be absolutely awful at work on OS X.

Safari is the only logical choice on OS X laptops, at least. Everything else just eats up the battery like crazy.

alexbecker2• 10 years ago

I've actually had the opposite experience. I work on Firefox support for a very heavy Angular2 app, and I usually see 2-3x faster reflows on OS X (and Windows) than on Linux, and faster paints too (although I haven't taken exact numbers, since reflows are the bottleneck). It seems that at least the graphics pipeline is much better optimized for OS X/Windows than Linux.

mnarayan012• 10 years ago

Firebug is quite often the culprit in these kinds of things in my experience. I can't live without it so I put up with it; if you're switching from Chrome however, it seems that you'd be better off just using the built-in inspection.

cpeterso2• 10 years ago

Yes. I don't know if this is still the case, but Firebug used to disable SpiderMonkey's JIT, even when not using Firebug:

https://groups.google.com/forum/#!topic/firebug/q2fuUraVDk8

mintplant6• 10 years ago

Have you given the built-in Firefox dev tools a try? If so, what did you feel was missing?

berberous• 10 years ago

This seems atypical. Given that you don't use any add-ons, consider resetting your profile. Alternatively, you may wish to disable multi-process for the time being if it's enabled for you (I'm sure there's a flag you can set for that if you Google for it), or upgrade to the beta or alpha builds. In my experience, the beta/alpha builds are usually plenty stable to be a daily driver.

dxxvi2• 10 years ago

I second that. My FF on Linux acted weirdly, then I removed the whole .mozilla directory (remembered to save the greasemonkey scripts I wrote, but forgot to save the favorite links :( ) and everything worked again.

robin_reala3• 10 years ago

To reset go to about:support and click on the button in the top right. A prompt will appear telling you what it’s going to do.

CaptSpify2• 10 years ago

For me, disabling JS by default was the culprit. I only turn it on for sites that actually need it, and it seems to work much better. YMMV of course

theandrewbailey2• 10 years ago

NoScript allows you to turn it off/on by domain.

cmurf• 10 years ago

It's tedious but I'd do some basic troubleshooting like embargo all addons, not merely disable, and trash the user prefs. If you use sync with a Firefox account, it'll have your history and bookmarks. Then add back add ons one at a time. I'm on 48.0.1 on Linux (on a Mac!) and I'm not having the problems you're describing.

0xfeba2• 10 years ago

Funny, I have the same lockups with Chrome. Middle-clicking a new link (to open in a new tab) freezes the entire browser (all tabs) until the new tab is done loading. I believe it's a extension problem (uBlock Origin, EFF Privacy Badger), but I don't really want to get rid of any of them.

dxxvi2• 10 years ago

I know this is not about ad blockers, but does anybody have any experience of using something like https://github.com/StevenBlack/hosts/blob/master/hosts instead of uBlock Origin? Or even use it in the hosts file of our dd-wrt/openwrt routers?

nfriedly• 10 years ago

No add-ins (except for Firebug).

You might want to try Firefox's built-in Web Console. I switched a year or two ago when it was less powerful than Firebug, but notably more stable, and I haven't gone back since.

except for an annoying OS-X multi-screen bug where it screws up the sizing

Can't argue with you there. At last exiting full-screen video no longer takes the entire browser out of full-screen.

mrweasel2• 10 years ago

You might want to try Firefox's built-in Web Console

Maybe it's a personal thing, but I much prefer the built in developer tools on Firefox to that of Chrome.

scrollaway3• 10 years ago

Firefox is frustratingly slow for me. I can't use it, have to stick to Chrome. Chrome which, by the way, crashes after 5 minutes of using video Hangouts and brings my 16 gigabytes ram PC to a halt when I open too many youtube tabs.

But firefox scrolling is especially slow. irccloud.com is the main culprit (and I use irccloud a lot). It's pretty bad in gmail as well.

The react app my team is working on is also far slower in Firefox. Try scrolling this in Firefox and Chrome and see the difference (the scrubber is at the bottom of the blue area, like a video). Butter smooth in Chrome, horribly slow in Firefox: https://hsreplay.net/replay/7VAKLeMNaXvshAawnUUoni

_dark_matter_• 10 years ago

This is super surprising to me, because irccloud is exactly what we use at Mozilla. Everyone seems happy with it's performance, and obviously most people are using FF.

bzbarsky4• 10 years ago

That last link looks about the same to me in Chrome dev and Firefox nightly, for what it's worth.

mstange2• 10 years ago

Scrolling especially should be much better in multi-process Firefox, because it will be async, so web pages no longer block it.

It doesn't fix performance in your react app, though. I've filed a bug about it: https://bugzilla.mozilla.org/show_bug.cgi?id=1300848

fbender5• 10 years ago

Do you have the latest graphics drivers? Some are really buggy and may be (one of) the cause(s).

bitL2• 10 years ago

Chrome became really really really slow on Ubuntu in the past year, no extensions installed. Every time I use it on a high-end Core M ultrabook, the CPU/RAM is eaten, the whole system becomes laggy, Google Hangout/Slack voice chat/Skype gets interrupted all the time etc. Not sure if you gain anything by moving back, sadly :(

bemused• 10 years ago

chrome on linux doesn't seem to get much attention for quite a while now: got affected by this [1] more than 2 years ago and been quite happy with firefox for surfing & dev ever since

[1] https://bugs.chromium.org/p/chromium/issues/detail?id=393395...

tdkl3• 10 years ago

but I was trying to reduce my Google dependencies.

Get Chromium.

http://chromium.woolyss.com/

aphextron• 10 years ago

Chromium would be great if it had a decent library of extensions like Firefox. As is, I'll stick with the latter.

effie2• 10 years ago

Also make sure you've got the Google services turned off in Settings/Privacy.

FT_intern• 10 years ago

Firefox has been painfully slower than Chrome by great magnitudes for me on a cheap Dell Windows laptop. I have almost identical extensions except for NoScript which might be the culprit, although it is also the main reason why I use FF over Chrome (and for enabled extensions on private mode, which defeats the purpose of private mode but it has been used more as a "don't save history" mode than private mode if you know what I mean).

ktta• 10 years ago

you can enable extensions for incognito in chrome too.

fbender5• 10 years ago

Did you try updating your graphics drivers? Sometimes bad drivers cause sluggishness.

wvenable2• 10 years ago

I've been a long time Firefox user (due to multi-row tab extension -- can't live without it). I can easily have 120 tabs open when doing research. Firefox has always been relatively performant even with that huge number of tabs visible and I almost never have to close it.

However, something recently has changed. Firefox will work fine and then something will cause it use a large amount of CPU that it doesn't let go of even after closing all the tabs. I have to close the whole browser and re-open it. This can happen a few times a day.

I'm hoping it's a regression that will be fixed soon.

aluhut3• 10 years ago

I've been a long time Firefox user (due to multi-row tab extension -- can't live without it).

Which one do you use? There are several.

charlee2• 10 years ago

I have this exact same problem, in fact I could have written your comment word-for-word. I also cannot reproduce it reliably, but it started in August.

I also switched from Chrome, but for slightly different reasons...I got to thinking about Google, and Chrome,and ad-blocking, and it occurred to me that Chrome will probably never allow extensions on Android, not because it's technically not possible, but because ad-blocking is bad for business (if your business is showing ads).

How does the Chrome team reconcile this? We know that ads are annoying, and ads can track you, and some ads can even be malicious. It seems that blocking ads make a browser better. But the Chrome team is not allowed to block ads on Android, so they are purposefully making a worse mobile browser.

Anyway, I do a lot of surfing at night in bed on my Android tablet, and I'm going to stick with Firefox.

tdkl3• 10 years ago

Based on CAF Chromium, with integrated ad-blocker and DuckDuckGo as an search option : https://play.google.com/store/apps/details?id=com.mokee.yubr...

nl3• 10 years ago

Just an update - I did the "refresh profile" thing suggested here, and I've been 2 days without lock-ups.

cesnja2• 10 years ago

I occasionally experience similar slowness. In my case, it's caused by bloated sqlite database. After I delete all history, the browser is snappy again.

fbender5• 10 years ago

Mind filing a bug with specifics at bugzilla.mozilla.org? There's probably some low-hanging fruit there …

dragonne• 10 years ago

It is frustrating that Firefox makes it so easy to delete recent history, but difficult to remove old stuff. I really only want the last month or so; why do I have to clear it manually?

fbender5• 10 years ago

I found ads to be a major hog (and some bad scrolling scripts on popular or design-y sites). I'm using uBlock Origin or Adblock Plus, which gives a great responsiveness boost for far too many websites.

ci5er• 10 years ago

I can't find any other reports of it, but I was on FF 47.0.1 on Windows 7 and upgraded to 48.0.1 (For some reason I never upgrade when prompted on the XX.0 releases). And I would have intermittent whole-browser freezes that lasted about a minute. Not like one every 4 hours, but more like once every 5~10 minutes. Back down to 47.0.1.

I tried again on 48.0.2, ... same thing.

I can't nail it down specifically enough to file a proper issue report, so I guess I'll keep trying when the next releases come out.

I also notice that webapps seem to like Chrome better, but FF usually performs "well enough" that I've put up with it because I find it (for me) to be markedly better at helping me deal with the ginormous amounts of text content I apparently consume. But webapps have been degraded enough for the last month-or-so that I'm running chrome in parallel just for those.

I love FF and I'm hoping that this is one of those isolated things that the fabulous team over there fixes before too long. (I suspect it's a JS related something or other, but I haven't been able to track it down specifically.)

(Wow. Now there is a rambling comment!)

kuu• 10 years ago

I have the same problem in Windows 10, since a month or two (I think since v48) something seems to be broken... Sometimes the browser will freeze or some sites take too long to load properly...

mschwaig• 10 years ago

Did you try updating your graphics driver or disabling hardware acceleration (about:preferences#advanced)?

I had an issue but on Windows 10 where sometimes on video playback FF would just lock up and windows would restart the graphics driver. I also couldn't reliably reproduce it, it sounds like a similar problem.

This is my bug report: https://bugzilla.mozilla.org/show_bug.cgi?id=1280706

drzaiusapelord• 10 years ago

Its incredible multi-process took this long. Just goes to show you that your architecture decisions last a long time and are often difficult to change. Chrome had this from day one and never had a big and old codebase to worry about. Yet it took Firefox many years to get multi-process going and my understanding is that its much more limited and simpler than what Chrome or Edge do.

I'm also a little surprised there hasn't been an attempt to launch a completely new Firefox from the ground up. Regardless of what they're doing right now, its still a legacy code monster and much more laggy than the competition. Maybe this is Servo's ultimate purpose, but every Firefox advance is welcomed but always feels like another layer of lipstick on this pig.

Disclaimer: I use Firefox as my main 'non-work' browser several hours a day. Its good, but its very obvious when I'm not in Chrome from a performance/stability perspective.

jayflux3• 10 years ago

|| I'm also a little surprised there hasn't been an attempt to launch a completely new Firefox from the ground up.

I (hopefully) think this is what Servo is and will end up being. Has no legacy code, and is tiny in comparison to Gecko. It's OS support is fairly modern, and has no interest in supporting Windows XP. It's also written in such a way that allows them to be more multi-threaded and more concurrent than traditional rendering engines.

At some point you have to draw a line in the sand and start again, and I hope Servo is that.

You can check out https://github.com/browserhtml/browserhtml which is servo running standalone

Manishearth3• 10 years ago

I (hopefully) think this is what Servo is and will end up being.

Not really the plan. Servo may eventually become a product, but that would be in the very far future. But Gecko can use lessons learned from Servo, and try to share code with it, so you can incrementally replace parts of Gecko.

and has no interest in supporting Windows XP.

I don't think that's the case? We don't have users on XP so we may not support it right now, but that's just an issue of priorities and not having the resources to fix all the things at once.

There are more XP users of Firefox than there are on Linux (probably for other browsers too). If Servo was a product it would probably care about XP.

williadc• 10 years ago

I'm also a little surprised there hasn't been an attempt to launch a completely new Firefox from the ground up.

Check out Servo: https://servo.org/

TheCoelacanth• 10 years ago

Servo is "just" a new rendering engine, though, not a whole browser.

radiantmilk• 10 years ago

As a fully rewritten browser engine, how will they handle malformed HTML? Will it, like other browsers, try to 'understand' and fix some errors or will it stick to the specification?

the84728• 10 years ago

Addons are a big part of the firefox ecosystem and they reach deep into the internals of the browser. So just replacing the browser wholesale would break a lot of things over the course of a single release.

kalleboo5• 10 years ago

They are deprecating the old add-on API now in favor of a more Chrome/Safari-like one now though

stcredzero2• 10 years ago

Its incredible multi-process took this long. Just goes to show you that your architecture decisions last a long time and are often difficult to change

The same happened with the original Mac OS!

kalleboo5• 10 years ago

I'm also a little surprised there hasn't been an attempt to launch a completely new Firefox from the ground up.

It's interesting how the history goes.

We had Netscape 4, and it was crap (remember when resizing a window reloaded the whole page?)

So they was a ground-up rewrite of the rendering engine, resulting in Gecko, which was put in Mozilla.

Gecko was great but Mozilla was a bloaty amalgamation of features, so this was created a pared-down Gecko version called...

~Phoenix~ ~Firebird~ Firefox, which had the great rendering engine and a lean, native-like UI.

Then there was KHTML/KJS which was built for KDE, which had a lean architecture but didn't have the investment to get the compatibility 100% there...

Which Apple then poured into the project in their fork to WebKit, which paid off in spades on the resource-limited iPhone a couple years later.

But Safari was only on the Mac (aside from their "Cocoa on Windows" version and a few open source ports), which Google took as a reason to create...

Google Chrome, where the biggest innovation was rendering each page in its own process (since Safari could freeze up due to one page going in an infinite loop in JavaScript).

So will someone out-Firefox Firefox and do what Chrome was to Safari for multi-processing? Take the rendering engine and put a new chrome on top of it?

flukus3• 10 years ago

A few additions:

KHTML was originally forked from QHtml (part of Qt) I believe (or they shared the same origin?).

Webkit later became a Qt component, the circle of life.

I would argue chromes biggest innovation was the UI, putting the address bar as part of the tab where it belongs (cue holy war) and better tabbing in general.

Mozilla was also the first to integrate the search and address fields, which they unfortunately stripped out of firefox (and many people think chrome introduced it).

bwat483• 10 years ago

So will someone out-Firefox Firefox and do what Chrome was to Safari for multi-processing? Take the rendering engine and put a new chrome on top of it?

I highly doubt it, as there's very little incentive to create a gecko based browser vs a webkit or chromium based browser.

faragon• 10 years ago

Why? Except because of memory allocation serialization impact, there is no reason for a multithreaded process being slower than multiprocess. And if that's the cause, it could be solved using multiple heaps.

gravypod• 10 years ago

I'd guess that it wasn't the effect of switching from threads to processes that made the difference. It might come from code clean up and fixing some of the IPC that was going on. This would have probably seen many eyes in the firefox dev community so they might have made some efforts to clean up the code under the hood before closing the lid on it.

But I might be wrong as I have no ties to the development process of this.

scott_s• 10 years ago

My guess is that the multithreaded-only design wasn't as good at exploiting the inherent per-tab parallelism. The original architecture probably had some synchronization points (perhaps related to the common GUI code) that did not allow processing each tab truly independently. Regarding memory allocation, yes, in theory one should be able to try to avoid sharing data across threads, which should eliminate serialization on memory allocation and deallocation. But I can easily imagine that in practice, in a codebase that was not explicitly designed to process each tab in parallel, you'll still end up sometimes sharing data across threads.

I imagine they could have instead changed the architecture to be multithreaded in such a way to avoid my imagined sequential bottlenecks, but performance is not the only goal in a multprocess architecture. Sandboxing is another design goal, and that cannot be fully achieved with a multithreaded design.

This is all speculation on my part, so if anyone who actually works on the Firefox codebase could correct me, that would be very welcomed.

mintplant6• 10 years ago

One near-future goal is to move the compositor out from the UI process into its own separate one. Mozilla currently maintains a sizeable blacklist of {feature, GPU, OS, driver} for hardware acceleration because glitchy drivers cause whole-process crashes. When this occurs a separate compositor process can be restarted instantly without the user noticing.

bzbarsky4• 10 years ago

The current setup is not multithreaded in the relevant ways.

So the real decision point was whether to take an existing large C++ codebase and try to safely make it multithreaded or to take that same existing codebase, run it in separate processes, and make the communication work.

The latter is actually a simpler engineering task, in my opinion, and paves the way for doing things like sandboxing and whatnot once the initial rollout is complete.

uuoc2• 10 years ago

This should come as no surprise. It has always been an epic architectural mistake to use the same single non-reentrant Javascript engine to both render the UI and run JS for webpages in Firefox. This change will finally undo that huge mistake made so very long ago.

addicted• 10 years ago

JavaScript was little more than a toy scripting language used to animate stuff and change layouts when html/Css wouldn't fit for the greatest part of firefox's early existence. At the time it was probably the appropriate decision.

Now, it's a huge burden on Firefox and it's good to finally see progress on this front.

creshal4• 10 years ago

JavaScript was little more than a toy scripting language

No, back then Mozilla was trying to do the entire UI logic in JS. That was the whole point behind reusing the same JS engine for both: https://en.wikipedia.org/wiki/XUL

eatbitseveryday• 10 years ago

Since Javascript is just a language that happens to execute client-side within the browser, why can we not use an existing language syntax (or an existing language entirely) for this purpose? For example, Go, Python, Rust, or perhaps Perl? It feels like a broken shell scripting language that really should have been replaced.

the84728• 10 years ago

Multi-processing is an orthogonal issue to single-threaded JS.

The same asynchronous logic used to have multiple browser engines with multiple JS runtimes talk to each other could also be used for multi-threading while making shared, thread-safe datastructures of non-JS parts much easier.

bitL2• 10 years ago

Yeah, I never understood why they did it, it was obvious a single unresponsive tab would block the whole browser, but I guess fresh blood at Firefox needed to gain experience...

rsp1984• 10 years ago

“We can learn from the competition,” said Dotzler. “The way they implemented multi-process is RAM-intensive, it can get out of hand. We are learning from them and building an architecture that doesn’t eat all your RAM.”

That's the money quote here. I've been waiting for this for a long time actually. Every browser I've tried except Firefox just basically eats all my RAM and other app performance (e.g. compiling stuff) goes down the toilet.

Then on the other hand FF has not been so snappy and responsive traditionally. So responsive + soft on RAM is the combination I've really been waiting for. Let's hope they can deliver.

toxican• 10 years ago

A few years ago if you'd told me FF would be the RAM-sensitive browser, I'd have laughed you off the internet.

AdmiralAsshat• 10 years ago

And they're still at it, too: https://areweslimyet.com/

marcosdumay2• 10 years ago

Funniest thing is that FF memory usage didn't go down that much. For me it is even higher.

zobzu4• 10 years ago

honestly its been years now that FF is one of - if not the - browser which uses the least amount of RAM.

Its also one of the most energy efficient on most platforms.. (edge is more efficient on Windows, though it does not take into account that the engine is always running even if another browser is in use)

flukus3• 10 years ago

A few years ago I had a a ptoato android tablet. The thing couldn't play music without stuttering, chrome would take more than a minute to render a page, then another minute if you scrolled.

Firefox was first released for android around the time I bought it though, and turned it into a capable web browsing device.

schoen• 10 years ago

Can someone describe what the "architecture that doesn't eat all your RAM" here is? Is it possible that it will inadvertently provide weaker security protections between tabs than the more naive and RAM-intensive architecture?

0xfeba2• 10 years ago

Duplicated memory and just overhead from running separate processes (Chrome), as opposed to shared memory and less overhead from a single process (IE? Old FF).

As for security, no, unless there's some unknown vulnerability now (or that they create), that will be ported over and somehow more effective between (potential) processes. So, I doubt it.

formula1• 10 years ago

One thing that firefox does well is suspend pages you are not using and likely saving the state of the appllication to the disk. Another thing can be for when multiple pages use the same or similar javascript/css/external respurce files, they only need to be loaded once into memory. Another thing canbe that certain javascript globals would only be loaded when the getter asks for them. Something like AudioContext and RTCPeerConnection are not widely used, by changing the name in the global space to a getter that loads the libraries instead of providing them on start may reduce load.

There are many things that can be done

the84728• 10 years ago

a) your security should not rely on process isolation. the javascript engine etc. should be secure in the first place. if you can break out of the JS sandbox you can already attack addon scripts (think password managers)

b) hosting some tabs in the same, sandboxed process is still stronger than hosting all of them in the parent process.

wingless• 10 years ago

Yes, that is definitely possible. Chrome isolates data as much as possible between processes. A shared cache would be a big improvement in memory usage, but security is the biggest concern.

tomp• 10 years ago

Nevertheless, I prefer Chrome's approach. I can just open the process explorer and kill a few of the most memory hungry tabs/tab groups; with Firefox, there's no other solution than killing the whole browser.

Can't wait for this fix!

NoGravitas• 10 years ago

I believe in current Firefox you could kill the memory-hungry tabs, then go to about:memory and force a GC. Or restart and let your tabs stay open but not loaded until you use them. Not that either of these is super-convenient.

kohanz• 10 years ago

I was in the same situation as you and found that in the end Safari fit that bill for me (not sure if it is an option on your platform). Responsiveness is probably somewhere between Chrome and FF, but memory usage has been much better.

Angostura• 10 years ago

If you're on Mac, Safari is good in terms of both memory and cpu.

r00fus• 10 years ago

Safari is great, but doesn't have a great solution for vertical or tree-style tabs, which are heavily used in my workflow.

FF is for work, Safari for non-work. Chrome when I need Flash or testing Chrome-only content (some great plugins) - pretty rare.

rimantas• 10 years ago

And battery life.

newscracker3• 10 years ago

Safari is my favorite on the Mac since it's quite lean (memory wise, CPU wise and energy consumption wise) and responsive. But I miss extensions from Firefox and don't use Safari if I have to work with several tabs and windows. There are several little things that make using Firefox much better for my experience.

Firefox is my favorite browser overall (especially on other OSes), considering its smaller memory footprint and lack of sluggishness with tons of tabs (compared to Chrome) and the fact that it's from Mozilla, a company I hugely respect and admire (even with some of the missteps that have happened in some aspects in the past few years).

hokkos• 10 years ago

Webkit also learned from the experience of chrome with WebKit2 architecture, and uses less ram https://trac.webkit.org/wiki/WebKit2

chakalakasp• 10 years ago

The thing is, workstation and laptop RAM is so cheap. If your browser is eating an appreciable chunk of your 16 or 32GB of RAM, you need to close some tabs.

watwatwatwat• 10 years ago

I'm tired of this argument. My browser is not the only piece of software I'm running. I usually have virtualbox, a compiler, an IDE running. And sometimes some scripts/programs I wrote that does intensive computing on some data set.

fnj• 10 years ago

Forgive me; your statement does not make sense. First you're implying that all you have to do is get more RAM; then you're admitting that you only have a certain amount, so you should stop using so much.

In actuality, very few non-server PCs can be populated with any more than 16-32 GB. It's not a matter of bother or expense at all. It's a design limit.

Every single time I've driven a non-server PC into hopeless swap thrashing, it's been the damn browser that was the culprit. Don't bother trying to advise me to bookmark 100 open tabs in 10 windows to save space. That is (forgive me) an insult to one's intelligence.

smsm426• 10 years ago

No, it's not cheap at all, because buying the physical chips is not the end of it. For some laptops - like macbook air - there is a limit of how much they support, and it's not that high. Some models are not even upgradable - they just have soldered-in 8G and that's it.

you need to close some tabs

This comment is really not contributing anything to the matter. Obviously everybody could think of closing tabs without you advising to do so, and if they don't, it may be because they want these tabs around. I regularly keep a lot of tabs around because this is my work pattern.

minitech• 10 years ago

Student here. Can't afford "cheap" laptop RAM.

blakeyrat• 10 years ago

I'm not a Firefox user, but considering that 95% of the time about half my RAM is going unused (well, Windows uses it for pre-caching random stuff it thinks I might need), I don't mind high memory usage as long as I get a benefit for it.

If it's just inefficient, that's a different story.

xorblurb5• 10 years ago

The thing is HD caching is extremely elastic. Memory consumption by a browser, less so (maybe even nearly not at all elastic).

So what you might think of a benefit one day might be a drawback the next if you want to open twice the same amount of tabs, or run a big build in the background.

Lagged2Death2• 10 years ago

I'm surprised by the responsiveness claims because my ancient Windows PC runs FF superbly. I think Win10+FF is already more responsive now than it has ever been.

I am also using Ghostery, however, with virtually everything blocked. It wasn't my intent to block ads, but that's mostly how it works out.

jayflux3• 10 years ago

It depends on operating system. For instance on OSX Yosemite Firefox is sluggish, yet on Windows 10 it's fast. People should really mention what operating system they're running when they say "x is slow"

shmerl2• 10 years ago

A pity it's now practically useless, since many add-ons need modification to keep electrolysis enabled. In a year or so things will be better I hope.

the84728• 10 years ago

Only if addon authors fix them. Afaik mozilla devs aren't planning on adding additional compatibility shims.

shmerl2• 10 years ago

Well, if add-ons are maintained, authors will fix them. Unmaintained add-ons is a bad deal anyway.

mintplant6• 10 years ago

There's currently a compatibility shim in place, however using it can lead to poor performance.

Kenji• 10 years ago

Does that mean my website will now fit on a 700% narrower screen?

rl3• 10 years ago

This will probably have significant implications for web worker behavior.

If memory serves, workers are unable to directly communicate with each other in Firefox when the UI thread is blocked, because message handling runs on the same process as the UI thread. Chrome doesn't have that limitation since it's already a multi-process architecture.

spynxic• 10 years ago

Wonder if Graal.JS could offer any improvements to the browser

AstralStorm6• 10 years ago

How much extra memory is now used by it? All this improvement is meaningless if it hits swap on a typical machine.

asadotzler• 10 years ago

We saw about 20% overhead going from one process to two.

nchelluri2• 10 years ago

Ah, so because I use addons I might not be seeing these benefits for some time? Good to know.

Chrome is right now way faster for me, but I prefer using FF so I am in the Mozilla camp right now. Eager to see this rolled out. Early versions of e10s were very fragile and not so usable for me, so I switched it off.

semicolon_storm• 10 years ago

I use quite a few add-ons and my e10 was disabled because of them. I force enabled e10 using Mozilla's guide https://wiki.mozilla.org/Electrolysis#Force_Enable and didn't notice anything breaking (yet).

nchelluri2• 10 years ago

thanks, I've turned it on; I'll see how it goes.

Sylos6• 10 years ago

It's usually less so things breaking and rather just being slow. For example, if you use NoScript, it often takes really long to load up a page, because that's when NoScript becomes active, and because NoScript isn't yet properly e10s-compatible, it just forces Firefox to do things in a single-process way. But that actually makes things slower than when Firefox is running without e10s.

about:performance is a rather good indicator, if an add-on is causing slow-downs.

zyngaro• 10 years ago

On which OS? On OSX I almost gave up on ff because it frequently freezes for no apparent reason.

Animats• 10 years ago

At last, you'll be able to view other pages while Facebook's Javascript is in an infinite loop.

With thousands of MIPS and a gigabyte of memory on the job, it had better be responsive. It's pathetic how much compute power it takes to run a browser. Even when the pages aren't doing anything interesting. It's not like they're running a 3D game or something.

midnitewarrior• 10 years ago

Why is there still a browser capable of having its performance improve by 400-700%? This browser is a decade old. It is clearly the CPU and RAM hog everyone has said its been.

Sylos6• 10 years ago

Responsiveness is not the same as performance. Responsiveness is how quickly it responds to user-input. So, for example, if you click on the Back-Button, then it highlights the Back-button to let you know that you have indeed successfully clicked the Back-button and that it is working on going back a page. How quickly it highlights that button after you've clicked it, that's how responsive it is in that aspect. Actually going back a page, i.e. loading the page from cache and rendering it, that's then again a performance-metric.

But you are also quite clearly underestimating how terrible modern browsers are. Mozilla is also currently working on a research browser called "Servo", which has been written completely from scratch and employs a more modern architecture, and that isn't too far away from 400-700% performance increase over other browsers in certain aspects.

corford2• 10 years ago

I've got 732 tabs open at the moment and performance is still fine. The upper limit for my machine seems to be about 745 ish - much beyond that and FF runs out of video memory or something (new tabs render sites half black).

ActsJuvenile• 10 years ago

You need to get a better handle on your life.

j3097736• 10 years ago

Some bugs might only occur on extreme conditions.

corford2• 10 years ago

Haha yeah :) TBH I have no idea what half of them are any more. I've just never gotten around to pruning them since FF works well enough and load up time isn't too bad (< 12 secs).

brokenmachine8• 10 years ago

Lol, I've seen that issue myself, but only at maybe half that (ridiculous!) number of tabs.

If you make the window smaller you can still do some stuff, but you need to restart FF (and wait for your 732 tabs to reopen!) to get everything back to normal.

Computers are meant to solve problems, not create mindless tab-closing jobs for humans! We should never have to close a tab! Tab hoarders unite!!

matt_wulfeck• 10 years ago

I don't know. "Eating up all my ram" seems perfectly acceptable to me at first glance. I purchase ram and use it in my computer for expressly this purpose: to improve the responsiveness and snappiness of programs.

Where do you draw the line at what's acceptable use of ram and what's not? I have 32G on my desktop, is it unacceptable that chrome is using 8G?

scott_karana• 10 years ago

I think it's "inefficient use of RAM" that they're talking about.

If Firefox should conceivably run in 8GB, but ends up trying to use 33GB (in your case), that would obviously be bad.

Mozilla just targets efficiencies that include people with much, much less RAM than you for their baseline.

Sylos6• 10 years ago

The question is how sensibly Chrome uses those 8 GB of RAM. We are very much still at the point where (most likely) not your entire hard drive + whatever data is generated on the fly or loaded from a webpage fits into your RAM. As such the often repeated statement that "unused RAM is wasted RAM" is still nonsense. There is always something which could be cached in RAM and would provide a speed-up of some form. If Chrome reserves 8 GB of that for itself, then it should utilize that in some form, so that it's used more sensibly than it could be used by other programs.

For example, Chrome could keep a hundred pages from your browsing history cached in RAM, so that when you go back, it loads that instantly. That is a speed-up, but you probably will never go back one hundred pages in a row. So, another program could almost certainly use the RAM more sensibly than Chrome, if that was the case.

And what Dotzler is referring to, is the worst kind of offense. If his promise holds true, Firefox's implementation will be just as good as Chrome's in terms of speed, but will use less RAM. Which would mean that Chrome is using some RAM without providing a speed-up, i.e. actually wasting that RAM. And that is always unacceptable.

Zpalmtree• 10 years ago

I hope this works out well, I often see my firefox freeze due to heavy extensions, and would much prefer if it just froze one tab. For me, higher ram usage wouldn't bother me if it meant no freezing, but I imagine it does matter to most of the audience, and it's too much effort to develop two feature sets.

zobzu4• 10 years ago

Ive been testing Firefox e10s as in multiprocess Firefox for months now (in Nightly, and even before that).

At some point it became very usable and it still is today. HOWEVER while the interface is, yes, much more responsive, it is also NOTICEABLY slower.

I ended up reverting recently and I'm using single process Firefox right now. Its fast even thus from time to time the UI may block if there's heavy stuff going on.

Whatever makes e10s slow they gotta fix it... I suspect there's a lot of synchronization code.

wcummings• 10 years ago

I've been running Firefox Developer Edition (which has multiprocess) for a few weeks now, it's noticeably faster for most sites.

eberkund• 10 years ago

You can install v51 nightlies from here: https://nightly.mozilla.org/

Double_Cast• 10 years ago

I use both FF and Chrome. Chrome is more heavy-duty, so I use it for Dedicated Browsing. I've set it such that when something calls me away and I close Chrome, the tabs will reopen so I can continue where I left off. But sometimes I'll be doing something in another program and need to look up something quickly. In such cases, I don't wanna wait while Chrome reopens 11 tabs so I'll use FF instead. My FF-tabs aren't saved between sessions, so I only see a single fresh tab when I open it.

Hydraulix989• 10 years ago

Is this different than multithreaded? What are the benefits? Is it really just about isolation/sandboxing?

Why so much RAM? I thought the OS could share pages across processes...

dschiptsov• 10 years ago

So, finally, some emerging consensus among packers that pthreads is a brain-dead concept, and thousands of its advocates are just narcissistic idiots, loving to hear their voices?

sambe• 10 years ago

Every odd browser performance topic is dominated by people claiming Firefox has been faster and more scalable than Chrome for years. Every even one is dominated by people claiming the opposite.

This thread seems to be the latter, my experience is the former. If you see bad performance/memory usage with a small number of tabs, are you sure that it's not due to bad plugins?

I think about:memory and about:performance could be helpful. I seem to recall they used to have some reporting on known problem plugins (perhaps as part of https://www.mozilla.org/en-GB/plugincheck/).

Multi-process Firefox brings 400-700% improvement in responsiveness

Related Stories