Firefox 22 Released, Boosts 3-D Gaming and Video Calls

Today Mozilla announced the launch of Firefox 22 for desktops and Android devices. For the desktop version, WebRTC, the open source browser-based communications API, is now enabled by default. "This technology makes it possible to place and receive video calls from a mobile or desktop browser or share live video, files and images with friends and family." Firefox 22 also has support for the asm.js subset of JavaScript, which allows for big performance boosts on graphically complex applications in the browser. (We saw a demonstration of this a while back.) Other new features include display scaling options for making text bigger on high-res displays, better WebGL rendering performance, word wrapping for text files displayed in the browser, and the ability to change the playback rate of HTML5 audio and video. The new Android version features include tablet UI support for smaller tablets, and a fix for scrolling in nested frames.

Comments

So...

[Jmc23]

when can I use it to run emacs?

Re:So...

[KiloByte]

Do you prefer a slow emulated real system, or a fast but minimal toy?.

Re:So...

[Jmc23]

I realize now that's probably too many meta levels for most people.

Why would a browser be an operating system?

Re:So...

[K. S. Kyosuke]

Well, an elisp implementation for JavaScript shouldn't be all that difficult to write. ;-)

Re:So...

[Lennie]

If not, you can use Emscripten/asm.js to just recompile the runtime as Javascript and be done quicker.

Re:So...

[K. S. Kyosuke]

That would be terrible, performance wise. For example, compiling a garbage collector in C to JS (even with asm.js) to run on a byte array can't possibly be better than using native JS objects as representation of ELisp objects, what with all the effort invested by Mozilla and Google into high-performance, incremental, generational low-pause collectors written in C++.

Re:So...

[Lennie]

Well, it depends.

This is the text of a slide of the presentation linked below:

"A Unified Approach?
Should we compile entire VMs from C/C++ to JavaScript, and implement JavaScript-emitting JITs?
Seems the only way to run most languages with perfect semantics + maximum speed
This is why I believe C/C++ to JavaScript translation is the core issue regarding compilation to JavaScript"

http://www.ustream.tv/recorded/29324270

http://kripken.github.io/mloc_emscripten_talk/

Re:So...

[K. S. Kyosuke]

Seems the only way to run most languages with perfect semantics + maximum speed

You'd first have to have an equivalent-quality generational collector in the original C code to reach the Javascript GC performance level - AND, in addition, the C++-to-JS compiler would have to generate code with zero drop in performance compared to native code. I don't think that the original Emacs elisp GC is *that* good.

Also, after compiling the Emacs elisp VM from C to JS, the only thing you get is a bytecode-interpreting VM (with a lousy GC to boot), whereas a Javascript implementation of the VM could arguably generate JS code on the fly that would take advantage of the V8/JaegerMonkey/whatever engine's inlining and runtime type specialization features, offsetting any possible disadvantage. Really, I'd be surprised if a good elisp VM written in Javascript weren't faster than the original elisp VM from GNU Emacs compiled from C to JS (even asm.js, for that matter).

Re:So...

[cyfer2000]

yes! http://bellard.org/jslinux/ wait for it to boot, then just type "emacs", have fun!

Re:So...

[haruchai]

JSUIX may be very fast but JSLINUX is NOT that slow
                                                                                                                   

Re:So...

[johnsnails]

When can I use emacs to run firefox FTFY

Re:So...

[Blaskowicz]

Worst name ever, why did they pick a name that is a homonym of an already existing browser? They could as well have named it Goggle Craum or Nestcape.

Re:So...

[Warbothong]

Ironically, JSUIX contains Vi but not Emacs ;)

Boosts 3-D Gaming and Video Calls

[nitehawk214]

Is there something named Firefox that isn't a browser but uses the same silly exponentially increasing versioning scheme?

Re:Boosts 3-D Gaming and Video Calls

[ntshma]

I was wondering this too.. Remember when Firefox as fast and lean?

Re:Boosts 3-D Gaming and Video Calls

[h4rr4r]

No, I do remember phoenix being that way though.

Re:Boosts 3-D Gaming and Video Calls

[gl4ss]

No, I do remember phoenix being that way though.

mod up. the slow downfall started when all the cruft doers at mozilla got on the phoenix train..

Re:Boosts 3-D Gaming and Video Calls

[LordLimecat]

It was also buggy and crashed a lot.

And really, it wasnt the browser but the web that was "fast and lean".

Re:Boosts 3-D Gaming and Video Calls

[MrBandersnatch]

Its still buggy and cra

Re:Boosts 3-D Gaming and Video Calls

[Merk42]

So... back before CSS3 and HTML5 existed?

Re:Boosts 3-D Gaming and Video Calls

[TangoMargarine]

Yes.

Re:Boosts 3-D Gaming and Video Calls

[PixetaledPikachu]

just think, we'll be on version 100 in five year. god damn, what is it with marketing departments and churning version numbers and where things are located. they need to die in a fire, slowly

nope, they will simply change the name to firefox GT, and restart the numbering scheme. Or invent a new one. Mozilla Firefox GT 7670s

Re:Boosts 3-D Gaming and Video Calls

[mspohr]

I believe that they have a linear arithmetic increasing number, not exponential.

Re:Boosts 3-D Gaming and Video Calls

[partyguerrilla]

Your ideas are intriguing to me and I wish to subscribe to your newsletter.

WhooHoo

[Nov8tr]

Now I'll be able to play games on my Mac!!

Fingers crossed for asm.js to take off

[Rubinhood]

I have big hopes for asm.js. Even with its teething problems, it's the best chance we have for a truly multi-platform common ground to develop networked apps in.

At the same time, this awesomeness has traditionally been ignored by the big players who desired fragmentation. Hopefully this time is different, as all browser vendors have a lot to lose if they are the last to implement asm.js.

The big missing feature is threading - here's hoping for an extension to asm.js to make it complete.

Re:Fingers crossed for asm.js to take off

[Lennie]

On the issue of threading:

That is what webworkers is for.

I know it wouldn't cover all the uses of threading, but it fit a lot of use cases.

Since we're all paranoid now, how secure is this?

[Bearhouse]

Not trolling, straight question. I know nothing about webRTC; are communications 'secure' by default?

Does it stop crap code ?

[Anonymous Coward]

setTimeout(function(){window.locationmanageQueryStringParam('source','autorefresh');}, 600000);

this bit of code is a nightmare on FF mobile, iam trying to read the comments and bam iam looking at the slashdot homepage ? WTF ? i didnt press back

sort it out slashdot, your code needs much more work and if you cared about the user you would NEVER reload a page the user didnt request.

Re:Does it stop crap code ?

[Inda]

It's a nightmare on the desktop too.

Open Slashdot link in new tab.
Read comments.
Close tab.

Look at Slashdot landing page scrolled to the top!

Now where was I? Dunno. How far do I have to scroll back down? Dunno. May as well go back to Reddit.

Sort it out Slashdot. You suck.

Re:Does it stop crap code ?

[jez9999]

I haven't had a problem. Then again, I do still have it set to use the classic discussion system. :-)

I want a car, no I want a plane...

[sinij]

I don't know all use cases, but I personally use Firefox to browse. Why do I want 3D gaming and video conferencing integrated into it? What next, preparing taxes?

Re:I want a car, no I want a plane...

[squiggleslash]

Because the web is increasingly an application delivery platform, and modern web standards reflect that. Many people may not like it particularly, but that's what it is.

Re:I want a car, no I want a plane...

[Anonymous Coward]

So that you can play web-based 3D games and video-chat on websites. This is a step forward that opens up a lot of new possibilities, but of course old people will whine about it. Stick with Lynx if that's your thing.

Re:I want a car, no I want a plane...

[MozeeToby]

I personally use Lynx to browse, why would I want images/audio/video built into it?

Re:I want a car, no I want a plane...

[washort]

"to browse" is a pretty vague term. This is equivalent to saying 30 years ago "I don't know all use cases, but I personally use my computer to compute. Why do I want 3D gaming and video conferencing integrated into it? What next, reading the news?"

The web browser is now a universal secure applications platform, standards-based, not controlled by a single owner. These are compelling reasons for people to want features added to it to compete with other applications platforms.

Re:I want a car, no I want a plane...

[0123456]

These are compelling reasons for people to want features added to it to compete with other applications platforms.

Thirty years ago, the old-timers used to call that 'bloat'.

I have absolutely no use for 3D gaming or video conferencing in my web brower, and I don't plan to be using it to run slow, ugly 'web apps'. Is there an option to turn all that crap off?

Re:I want a car, no I want a plane...

[archen]

Most of this stuff requires some javascript to function, so noscript is probably your only option.

Although in about:config you can set webgl.disabled to true if you want to disable that specifically.

Re:I want a car, no I want a plane...

[UltraZelda64]

Adding all this garbage is just setting the browser up to be like IE6 before it... a huge, bloated, buggy, major security risk. Most if not all of those things are already true to some extent, but at this rate it's only going to get worse. Once upon a time, a web browser just fetched web pages... now it's making it braindead easy to run unheard of amounts of potentially untrusted code. Beyore, you would have to download an executable in most cases or even buy a program at a store... now, all you have to do is browse a few web pages.

Re:I want a car, no I want a plane...

[Cyko_01]

So adobe flash can finally die

Re:I want a car, no I want a plane...

[0123456]

Do try to realize that YOU are not a representation of what everybody else wants or needs.

The people who claim everyone wants to run 3D games and video conferencing in their web browers might want to consider that comment.

I've certainly never, ever met anyone who does, though i'm sure they exist. I see no reason why this crap should be imposed on everyone just for a tiny few who think running Quake in Firefox is a really good idea.

Re:I want a car, no I want a plane...

[PurplePhase]

Awww, never played Spy Hunter in your Excel spreadsheet?

Re:I want a car, no I want a plane...

[SeaFox]

I don't know all use cases, but I personally use Firefox to browse. Why do I want 3D gaming and video conferencing integrated into it? What next, preparing taxes?

I do my taxes every year in Firefox... on the TurboTax website.

Re:I want a car, no I want a plane...

[Xest]

People don't like it because it requires fudge after fudge after fudge to make it work.

The problem is that everything from HTTP's stateless nature, through to Javascript's poor OO support mean that the underlying technologies aren't ideal for building larege applications. There's nothing inherently wrong with HTTP being stateless and there's nothing inherently wrong with Javascript having poor OOP support, but it does mean they're not ideal for building applications of a reasonable size. You can only fudge such technologies so far and this is why Google Apps as good as they are are still poor mans copies of MS Office/Open Office.

We've had things like Flash, we've had things like XBAPs but I suspect eventually it's going to have to be done properly - an open cross platform application development platform that possibly has it's own app:// protocol handler but that can hence cross-link too and from http based documents if we're going to really see the web advance. Either way, we're not going to be able to keep down the current path indefinitely, I can't see us being able to mangle http et. al. for another 20 years to suit changing requirements precisely because we are reaching the point where the required technology mangling is getting too prohibitive to do anything of any increasing complexity.

Languages like Javascript are what they are, not perfect, but not inherently bad, but it's not suited to large scale application development, it's simply the wrong tool for that job, and although you can fudge the wrong tool for the job into working for a while there reaches a point where this just stops being the case. At some point we're going to have to see a supplement to current web technologies built specifically for the purpose of application delivery and until it happens I don't really blame people for complaining about the current fudging that's going on to try and avoid that inevitable fact.

Re:I want a car, no I want a plane...

[squiggleslash]

I don't think anyone's arguing the web solution is perfect, but ultimately there is a massive demand for a distributed client server platform that's independent of any one vendor, and that's what's brought us here. That demand isn't just from "content providers" (I don't mean the RIAA/MPAA members here, I mean the wider group that includes everyone from bloggers to game developers), but also from users who are used to the basic web "visit a webpage and it just works" model and want that to be the way everything else works.

(BTW, not relevent to this discussion but FWIW I feel Javascript is actually a very underrated language, it's a shame its lowly beginnings and the fact we're still stuck as JS programmers interacting with the outside world using the DOM, are preventing it from being taken seriously. 99% of the time when I hear criticism it's either by people unaware of what it can do - hell, even Java's still catching up - or by people who confuse languages with APIs and whose beef is actually with the DOM.)

Re:I want a car, no I want a plane...

[tibman]

lol. Yes, you can do your taxes in the browser now.

Re:I want a car, no I want a plane...

[Xest]

I agree with you there's a good reason for the web as an applications platform, we just need better technologies to allow it to do that better because we currently risk breaking or at least over complicating the technologies that are used for static content in trying to make them something they're not.

I have no problem with what Javascript can do but I think even when you understand it's functional capabilities and so forth it's not a great language. The difference between declaring var and not causing the difference between local and global scope for example is braindead and it's handling of captured variables in closures is unhelpful. The point is that it's not that Javascript isn't powerful and that you can't do impressive things with it when you know it's quirks inside out but that there are plenty of languages that do everything Javascript does without those quirks proving that those quirks are demonstrative of original bad design than an inherent trait of Javascript's prototype/functional/oo hybrid design - the problem is it supports multiple paradigms but it supports them all in a sub-standard manner and that's the problem, none of the issues that make it sub-standard stop you doing what you need to do on a small scale but they do result in unnecessarily verbose and more awkward to debug code than is necessary such that as your code base grows maintainability of it gets pretty awful through no fault of the programmer but simply as a result of the language's quirks and half arsed implementations of various muddled paradigms.

So yeah, Javascript is powerful, yeah you can make it work, but a better thought through language would allow production of far better code again. Personally I think C# is a good example nowadays because it's a proper fully featured OO language from the outset, with fully fledged lambda support and optional DLR for dynamic typing or creating a prototype style environment where necessary. C# has evolved such that it's only added one paradigm at a time to make sure that as it has it does each properly, Javascript in contrast feels like everything was thrown in with no clear plan and not enough time spent ensuring it was done as well as it could be and that's the problem.

Boosts 3D Gaming and video calls?

[rbpOne]

That's two things i really never do within my browser.

Maybe in 23, they could boost performance and reduce memoryleaks instead of adding bloat.

Re:Boosts 3D Gaming and video calls?

[HaZardman27]

Have you thought that perhaps you aren't doing these things on your browser because they have been previously unsupported, slow, or lacking critical features of native clients? There was a time when videos weren't watched on a computer, because a computer was for performing math calculations. We do more things on a computer now because computers have gotten better at doing them. The same goes for web browsers, which are nice as a mostly standardized platform for running software on a variety of clients.

Re:Boosts 3D Gaming and video calls?

[sinij]

Have you thought that perhaps feature creep has bigger downside in context of browsers other than bigger footprint and wasted development time? Security for example.

Decades of constant battles with Java security should be abject lesson to anyone eager to swell functionality past any reason.

Consider how much better our browsing would be if Java never existed? I am eagerly await near future when marketing gets a hold of video conferencing and start throwing sales pitches at you or hackers figure out how to access video feed on your laptop.

Re:Boosts 3D Gaming and video calls?

[Cyko_01]

FYI: google chrome has this same "bloat" in development and stable builds as well as true bloat like NaCl - a brand new interpreted language on top of javascript that no one uses

Re:Boosts 3D Gaming and video calls?

[UltraZelda64]

I do VoIP in my browser, and I can tell you, while I like the capability to call people from my computer for free using Google Voice, I FUCKING HATE being forced to do it in a web browser. The Gmail site is a bloated pig, just like so many others these days, and Firefox itself is also bloated to hell these days. With 1GB memory, it is NOT a pleasant experience, and sometimes the damn plug-in even refuses to load. I literally cannot open Firefox with Gmail and Slashdot without the system swapping like a son of a bitch. If Google provided a "native" client, I wouldn't hesitate to use it and stop having to load and stay on a bloated web page within a bloated browser for the entire duration of the call. Simply put, 3D gaming and video calls are a DUMB idea for web browsers.

Re:Boosts 3D Gaming and video calls?

[rbpOne]

" There was a time when videos weren't watched on a computer, because a computer was for performing math calculations. We do more things on a computer now because computers have gotten better at doing them."

I play 3D games on my computer, but not in my browser. I watch videos both via browser and specific applications, but i dont use my browser as a Home Entertainment System.

Why would a browser do 3D-gaming and videocalls? Its nothing but bloat.

Re:Boosts 3D Gaming and video calls?

[Blaskowicz]

I find having a second browser is useful, such as Chromium, Midori or Epiphany - using the latter one lately. That way if you have to be stuck running a browser while you're out of ram and swapping you can quit or kill -9 the pig browser and still have some browser shit running in the secondary browser.

With 1GB, you ought to run a lightweight OS / environment (Windows XP, but it's deprecated, or LXDE, or some non xubuntu XFCE at best) and look for a memory upgrade unless you're on an old maxed out computer.

Re:Boosts 3D Gaming and video calls?

[UltraZelda64]

The problem with a lighter environment is that most of the bloat I mentioned is the culprit: the web browsers and the web sites. All it will do is allow me to load maybe one more semi-bloated page without swapping *quite* as badly.

Re:Boosts 3D Gaming and video calls?

[HaZardman27]

The point is that the term "web browser" is outdated. Today, web browsers are essentially compatibility platforms for running non-utility type software. The point is being able to create an application or game and allowing your users to decide what operating system they want to use, or whether they want to run your software on a desktop, tablet, or phone. It's possible to do those things natively, but it typically requires a lot of support and testing.

nabled by default

[rossdee]

" WebRTC, the open source browser-based communications API, is now enabled by default. "This technology makes it possible to place and receive video calls from a mobile or desktop browser or share live video, files and images with friends and family."

This doesn't sound very convenient - there are times that I am "browsing" when I don't want a video call suddenly interrupting me.

Luckily I don't have a webcam on this PC

Re:nabled by default

[washort]

Firefox handles it the same as geolocation or offline storage -- the browser asks you if you want to grant access to the camera and microphone.

Re:nabled by default

[washort]

what security concerns are you thinking of?

It handles the flying periodic table better

[sasparillascott]

Not perfect, but better... http://mrdoob.github.io/three.js/examples/css3d_periodictable.html

Re:It handles the flying periodic table better

[wonkey_monkey]

Under Firefox 21 I could drag and rotate quite smoothly, but the transitions from one model to the other were very slow. After the upgrade, it's all smooth.

Re:It handles the flying periodic table better

[TranquilVoid]

Same. 22 is noticably better than 21 especially in the layout transition. Further comparison;

Chromium - even faster and a little (but noticably) snappier when zooming or rotating
IE10 - like walking through quicksand. Still works though.

To be fair to Firefox, I have a literally hundreds of tabs open (most still unloaded) and Chromium has 5. For my use Firefox is still much preferred.

Re:How to disable?

[magic maverick]

Torbrowser, and you get the added benefit of Tor! Or, if you just want Firefox, download the latest ESR release (10.X I think). If you can find it.

Spot the odd one out

[wonkey_monkey]

WebRTC is now enabled by default.

Useful!

support for the asm.js subset of JavaScript

Impressive!

word wrapping for text files displayed in the browser

Decidedly underwhelming.

Re:Spot the odd one out

[partyguerrilla]

Scrolling horizontally is such a pleasant experience. What are these clowns thinking?

Re:Spot the odd one out

[wonkey_monkey]

If this doesn't get people flocking back to IE, nothing will.

And, just to chase some "informative" points:

about:config
plain_text.wrap_long_lines

Re:Spot the odd one out

[TangoMargarine]

Funny, my use cases are the exact inverse of yours. I have a JavaScript bookmark on my bar for exactly the last use case (well, for wrapping long <pre>s, actually). I don't have a webcam and don't intend to buy one, and would use different software for that anyway if I did, and have heard enough about Mozilla "focusing on speeding up JavaScript" for the last 12 releases that I'm sure it doesn't need more boosting for the time being.

How about *reducing* bloat? But yeah...

An OS to run "web applications"

[tepples]

The browser became an operating system when "dynamic web sites" became "web applications". The turning point as I see it came sometime close to when Firefox and Chrome added JIT recompilation for JavaScript in 2008.

Re:An OS to run "web applications"

[0123456]

The funny part is that, after they went to all that trouble, 'web apps' are now being replaced by plain old 'apps'.

Re:An OS to run "web applications"

[lgw]

Well, it turned out to be not a very good operating system, so it's losing to better ones. Not really a big surprise.

Re:An OS to run "web applications"

[tepples]

Well, it turned out to be not a very good operating system, so it's losing to better ones.

But you'd probably have to agree that the web is a much more widely deployed operating system than any single phone operating system, and the APIs of IE 10 Trident, Gecko, WebKit, and Blink are much more similar than the APIs of Windows, OS X, GNU/Linux, Android, iOS, and Windows Phone.

Re:An OS to run "web applications"

[lgw]

Sure, but unless you're coding from your basement, all that matters is that any given phone OS is big enough to justify porting your app to. And it may well be easier to write for your favorite platform and port to others than to write for a really bad platform once.

Most shops I know do both - they port to all the big platforms, plus support a stripped-down version to run as a web app (because you just can't do it right in the web app).

Re:Linux

[kthreadd]

It still does absolutely inane things like treating Gimp/Inkscape as valid a PDF reader.

Are you sure that's a Firefox issue? Here (Debian) Firefox picks up gimp as a PDF reader because the gimp package declares itself as being able to handle PDF.

Re:Linux

[MrBandersnatch]

I used to have issues with locking but it seems fixed with a Ubuntu 12.04/Gnome 3/FF 20 combo....

Are there still memory leaks?

[waterbear]

Is there something named Firefox that isn't a browser but uses the same silly exponentially increasing versioning scheme?

I used to use Firefox, and the thing about it that I remember 'increasing' was its memory consumption, due to leaks, up to the point that the computer practically froze. I stopped using Firefox when all I could get was denial that there were problems.

So how is the memory leakage issue now? (If there are reliable good reports, then maybe I might dare to try Firefox again?)

-wb-

Re:Are there still memory leaks?

[Lumpy]

Lot better. it is now far faster than Chrome. I have switched back.

Re:Are there still memory leaks?

[nitehawk214]

Lot better. it is now far faster than Chrome. I have switched back.

It can still leak memory and be fast. (well, right up to the point where the computer runs out of memory)

Re:Are there still memory leaks?

[nigelo]

Memory leaks are normally attributable to the plug-ins used, rather than Firefox, nowadays.

Unfortunately, memory leaks are usually blamed on the browser, not on a plug-in, regardless of the cause.

Re:Are there still memory leaks?

[rsborg]

Lot better. it is now far faster than Chrome. I have switched back.

It can still leak memory and be fast. (well, right up to the point where the computer runs out of memory)

Luckily modern operating systems are pre-emptively scheduled allowing you to simply kill and restart the offending App.

Are you by chance running Classic MacOS9?

Re:Are there still memory leaks?

[dsinc]

Funny you should say that. I regularly use Firefox and Chrome on two very different machines: one, is an anemic laptop with a Pentium T4200, the other a desktop with 8 cores and lots of memory. On the weaker machine, Chrome is visibly snappier - and never slows down at the end of the day. Firefox seems quite spry in the beginning, but quickly becomes visibly slower - not its rendering, but its general reaction, the awful, XUL-based interface. By the end of the day, right-clicking on a page and/or opening a new tab has a very visible, very annoying lag and the overall reactivity has decreased greatly. It's not a problem induced by some esoteric extensions (I only use Ghostery), I have enough memory (8GB, and the system memory load never goes beyond 50-52%), I only read text (lots of pages, though), no video, no games, flash is disabled via click-to-run. The faster/newer machine exhibits the same behaviour, it only takes a while longer for the lag to be apparent (due, obviously, to the increased computing power). I've always been a supporter of Firefox, I've been using it continuously since the 0.x era (in its Phoenix incarnation), but I'm not blind and statements like "FF is faster than Chrome" simply make no sense to me (and my browsing habits).

Re:Are there still memory leaks?

[TubeSteak]

Unfortunately, memory leaks are usually blamed on the browser, not on a plug-in, regardless of the cause.

Give me an easy way to trace which plug-in it is.
Surely Mozilla could do that?
They already tell me which plug-ins take a long time to load, why not some basic memory management?

Re:Are there still memory leaks?

[UnknownSoldier]

Haven't seen it in the twenty series, but after being a die-hard Firefox user since ~2000 finally switched to Chrome* last year. I suspect the video was the Flash plugin leaking memory. Firefox about:memory is a total joke. 2 GB to 4 GB usage after 30 days of heavy YouTube usage and the ONLY way is to close down and restart the browser.

* The reason: saw that Chrome was extremely fast (V8 compiles JS down to assembly) and has a beautiful "Tools > Task Manager" option that runs every web tab in a separate process, can precisely track memory usage per web page, and shows CPU usage and Frame Rate.

Re:Are there still memory leaks?

[Anonymous Coward]

about:memory / about:compartments

See https://developer.mozilla.org/en-US/docs/Zombie_compartments for a lot of the finer details.

If you just want it cleaned up, retaining bookmarks, history and passwords only use Firefox reset. https://support.mozilla.org/en-US/kb/reset-preferences-fix-problems

Re:Are there still memory leaks?

[Blaskowicz]

True that.
In my experience, Chrome is faster than Firefox, but uses too much RAM. Firefox "leaks" CPU and Chrome "leaks" memory - it doesn't really leak, it wastes memory by the ton because it was design that way. Sadly I don't have 8GB or 16GB so I can run Chrome or Chrome + Firefox, so I just run Firefox.

I run Epiphany on the side (lamely renamed "Web" by the Gnome team) which is fast, lean, almost has a really great layout, and is annoying like hell for not allowing to scroll the tab bar with the scroll wheel (I read the gtk bug reports, the feature was dropped from gtk in 2010 because one developer had a faulty scroll wheel and the devs convinced each others a scroll wheel confuses noobs)

Re:Are there still memory leaks?

[Inda]

I know!

I have an 'add-on' that fetches all linked images on a page and wraps them in <img> elements. One hundred 8-megapixel images on a page! No need to click anything!

My laptop, with 3gb of RAM, grinds to a halt. Hard disk thrashing; paging all those images to disk. Javascript engine moaning about taking too long.

What's a man to do? Fucking Firefox with it's bloated code and high revision numbers.

I jest (although the above is almost 100% true). I still think Firefox blows the competition away and I'm still suprised at the lack of love on Slashdot after all it did for us. Loading a HTML document with 100 large images is never a good idea. My add-on also does infinite scrolling and those 100 can become 1000 before I have to give up. I still remember the joy of downloading a single high res (1.3mp!) image nearly 20 years ago - how things have changed.

Plug-in fault? Firefox fault? Hardware fault? My fault for expecting too much?

Re:Are there still memory leaks?

[nigelo]

I'm pretty sure that closing a tab will not free up the memory in Firefox immediately due to the 'closed tab caching' (may not be the right jargon) that goes on in Firefox.

I'm talking about the ability to un-close a tab - right-click on any tab header and choose 'Undo close tab'.

Please feel free to correct me if I'm off-base here.

Re:Since we're all paranoid now, how secure is thi

[Anonymous Coward]

Not trolling, straight question. I know nothing about webRTC; are communications 'secure' by default?

They use (S)RTP for the transport:

http://www.webrtc.org/reference/architecture#TOC-RTP-Stack
http://en.wikipedia.org/wiki/Real-time_Transport_Protocol
http://en.wikipedia.org/wiki/Secure_Real-time_Transport_Protocol

The speicific protocol used is DTLS:

https://en.wikipedia.org/wiki/Datagram_Transport_Layer_Security

Piorities

[partyguerrilla]

They should focus on rewriting their engine and "boosting" memory management.

Re:Piorities

[Cyko_01]

They did a long time ago and are continuing to do so

Re:who cares ?

[0123456]

But fixing bugs is boring.

I like Firefox

[dskoll]

I know it's fashionable to give Firefox a hard time, but it's still my default browser after all these years. I use Firefox about 85% of the time and Google Chrome about 15%. Speed seems comparable, but overall the Firefox UI is still better for me.

Thank you, Mozilla developers.

For the Whiners

[caspy7]

For those who read the title and came here to moan about bloat:
The technologies mention in reference to 3D gaming are WebGL and asm.js. These serve to make things faster and their size is negligible (want to complain when the few extra bits in your JS engine make things go faster?). They can both be used in non-gaming situations, particularly processing-intensive stuff like dealing with images (processing, filters) and video (decoding - see ORBX.js). WebGL was already there, it's just better now.

You can disable it if you want, but WebRTC stuff doesn't load additional components (encoding/decoding video for instance) unless you're using them - which would be no worse than Flash (better actually). And just like with Geolocation, a site has to ask permission - to which you can say "never".

Chrome already has WebGL, WebRTC and is optimizing for asm.js. It's possible to land these without adversly impacting performance/responsiveness, and for the past year Mozilla has had their eye on the metrics.

An XMPP client?

[hobarrera]

So with all this new tech, it would be interesting to see a video-enabled web-based XMPP client sometime soon.
Does anyone have any now on something like this?

Re:An XMPP client?

[Anonymous Coward]

All of 5 seconds of googling;

http://blog.tmcnet.com/blog/tom-keating/voip/zoep-voip-client-with-firefox-plugin-and-jabber-support.asp

https://addons.mozilla.org/en-US/firefox/collections/bigou/xmpp/

support for asm.js

[Cyko_01]

Every browser, even IE, has support for the subset of javascript - that is the whole point! What is significant is that it has been OPTIMIZED for the asm.js subset

Any Downsides to Upgrading?

[PastTense]

Any Downsides to Upgrading?

For example are lots of extensions not working with the upgrade?

A Sad Fate

[runeghost]

For a project originally started to combat software bloat....

Chromebooks and Firefox phones

[tepples]

The funny part is that, after they went to all that trouble, 'web apps' are now being replaced by plain old 'apps'.

Unless you have a Chromebook or a Firefox phone. Their API for "apps" is the HTML DOM, just like the API for "web apps".

should be plugins

[Tom]

Time to move away from Firefox, it seems, and I've been using it since long before it had that name.

Why? Because it's going down the bloatware road that already destroyed Open Office. If you want 3D or video conferences, or kitchen sinks or coffee machines in your browser, a plugin is the proper way to go, period.

I don't want it. It want to display HTML pages. That includes Javascript and CSS and stuff, but why the f&%$! does it even have a plugin system if every newfangled crap gets thrown into the core code?

Firefox is taking up 650 MB right now. Yeah, that's with a bunch of tabs open, but still. 650 fucking MB to surf the web. It is by far the most memory-hungry app I have open all the time. The next in the list (Mail) clocks in at 160 MB, that's less than a quarter.

No, thank you. I don't want any additional crap in my browser.

Re:should be plugins

[elementalest]

I opened up a bunch of tabs (9 in total) in firefox 22.0 which came to a total of 398MB memory usage. I opened the SAME tabs in chrome and it came to 592MB. I opened the SAME tabs in IE10 and it was 354MB. I guess firefox is actually pretty good. Besides there are a stack of plugins that have been around for a long time that are not being integrated and probably never will. In any case IE10+ looks like a good option for you.

Video calls in a browser

[flimflammer]

So why exactly is this something we want in a browser? How much more bloat do we need before it's enough?

Video calls: Yes please.

[Trogre]

Anything to rid my friends and family's dependence on Skype(tm).

in other news...

[johnsnails]

FF23 is due in a few weeks.

not a big change

[robjacksond]

I prefer to Google chrome, but sometimes will switch to Firefox. The new features don’t have a big change for average users

I have bad news for you

[Hypotensive]

Go to http://jsperf.com/asm-js in Firefox 19: figures are the same for fast and slow.

Load the page in Firefox 22.0 (Mac OS X 10.7):

• fast: 6,539,184 ±0.48% ops/sec (74% slower)
• slow: 25,160,685 ±1.13% ops/sec (fastest)

They managed to optimise it to take four times as long to run the asm.js code in comparison to normal interpreted code. Great job!

Re:I have bad news for you

[tibman]

I get:
Testing in Firefox 21.0 32-bit on Windows Server 2008 R2 / 7 64-bit
fast - 39,535,593 ±1.52%fastest
slow - 29,803,623 ±0.78%24% slower

IE, Chrome, Visual Studio, SSMS and tons of other stuff open on a mid-grade laptop. BRB updating

Re:I have bad news for you

[tibman]

Lol, after upgrading to 22 it is worse : /

Testing in Firefox 22.0 32-bit on Windows Server 2008 R2 / 7 64-bit
fast - 5,678,041 ±0.35%84% slower
slow - 36,012,154 ±0.72%fastest

Re:Firemacs

[K. S. Kyosuke]

Really? Does this run Org Mode as well?

Bootstrapping a company

[tepples]

Most of your points sound reasonable except for one thing. I was unsure of the intended tone in one part of your comment. Is it bad to be "coding from your basement"? Is there a better way to earn the money needed to start a "proper" company?

Re:Bootstrapping a company

[lgw]

It's only bad if you're coding from your mother's basement. But if you're "not yet a company", I'd suggest writing for the easiest platform with a large user base, as the platform that lets you actually get a product out there the fastest is the best platform.