First Look At Firefox 3.0 Beta 2

DaMan writes "ZDNet takes Firefox 3.0 beta 2 for a spin and draws some conclusions that should be sweet music to Mozilla's ears. "Beta 2 feels snappier and far more responsive than beta 1 (or Firefox 2.0 for that matter) and I can feel the difference on all the systems that I've tried it on — from a lowly Sempron system to my quad-core monsters. No matter what you want doing — opening a new tab, moving tabs, opening up Find, zooming in and out of the page, bookmarking — it all happens swiftly and smoothly. What surprises me about the Firefox 3.0 beta is how many memory leaks that Mozilla have fixed. Complaints of memory leaks with Firefox 2.0 were met with an attitude of "Leaks? What leaks?" Considering that there have been more than 300 leaks plugged, it's obvious that past versions leaked like sieves.""

Comments

Hmmm...

[RotsiserMho]

But does it pass Acid 2?...

Re:Hmmm...

[HungSoLow]

You mean to say IE8 did, so you'd hope Firefox will!

Re:Hmmm...

[stony3k]

It is much more important to be compliant with CSS than just passing the Acid2 test, and so I really don't pay much attention to this test at all. There are better test suites out there, for instance http://www.css3.info/selectors-test/test.html.

We need to pay less attention to passing any one test and more to standards compliance as a whole.

Re:Hmmm...

[RotsiserMho]

Normally, I'd agree with you, but I think in this case it's different. It's all about public perception and to and extent, marketing. If IE8 can pass a test that's widely publicized and the latest FireFox can't, people may doubt that FireFox is superior. Of course people such as yourself will realize it doesn't mean much, but it's a very easy thing to point to and say "Hey it looks like Microsoft got something right."

Re:Hmmm...

[stony3k]

Even in that case, Firefox 3 beta1 was compliant and there seems to be a recent regression. So we can argue that Firefox was Acid2 compliant long before IE8. What Mozilla needs to ensure is that the final version of Fx3 is Acid2 compliant (which I have no doubt will be the case for exactly the reasons you state).

Re:Hmmm...

[ozmanjusri]

You mean to say IE8 did, so you'd hope Firefox will!

It's been claimed for IE8, but anyone can download the Firefox betas and check for themselves. Big difference.

Wouldn't be the first dose of vapourware to come out of Redmond....

Re:Hmmm...

[The_reformant]

To be honest its mostly irrelevant at this point since you're still going to have to support FF2, IE6 and IE7 for years yet.

Re:Hmmm...

[19thNervousBreakdown]

You're right. We won't be able to fix problems today, so fuck it, let's just never fix them.

Re:Hmmm...

[Anonymous Coward]

Please use the word "fuck" instead of "fscking". Who do you think you're fooling when try and alter the spelling?

Re:Hmmm...

[ozmanjusri]

I'd be more than surprised if that would be abandoned for the RTM version of IE 8 after announced on the IEBlog

Abandoned no.

Kludging to pass the test without actually implementing full CSS2 support, yes.

ACID2 is a test of a few of the hardest elements of CSS2, based on the assumption that if you passed the test, you'd have good support for the rest of the standard. If your goal was to just get the tick in the box for marketing purposes, it wouldn't be hard to just kludge it.

That's very much Microsoft's style. Look at how they're hacking ISO instead of fixing MSOOXML for a recent example.

Re:Hmmm...

[Michael Wardle]

Firefox is developing more quickly because it doesn't have all of the baggage.

Bullshit. Do you think Firefox doesn't have to render stuff written in Frontpage too? Mozilla pays just as much attention to quirks mode as Microsoft.

Memory Leaks?

[trytoguess]

I was under the impression that the issue was memory fragmentation. Ah well... does anyone have a link about this? I swear I read it somehere, or maybe it's from here heh.

Re:Memory Leaks?

[The MAZZTer]

It would take a really bad OS to make memory fragmentation a problem, since memory address pointers are virtualized (IE I'm talking about how process A can't access process B's memory and how the same numerical pointers in each point to different memory locations). Even Windows isn't that bad. Besides, the only performance metric any kind of fragmentation can really affect is speed, never size.

Or is this some misnomer or am I misunderstanding this?

Re:Memory Leaks?

[Frosty Piss]

Always the excuses come out, but never accepting the existence of a problem most "users" know exists...

Re:Memory Leaks?

[BZ]

What excuse, exactly? Being unable to reproduce a leak makes it hard to fix. Saying "I leak, but I won't tell you anything about how you could start looking for my leak" doesn't help get the leak fixed.

None of this is _good_. It's just a statement of fact.

Re:Memory Leaks?

[Kelson]

s/always/once/ (that I can remember)

And then Slashdot collectively declared it to be the official response, and repeated it over and over ad nauseum until people believed it. Kind of like the "Acid2 only tests error handling" misconception that came up several times earlier today, even though if you actually look at the description of the test, it's only one aspect among many.

Re:Memory Leaks?

[Seumas]

It's not bullshit. It's the response I have seen every single time the situation is discussed by anyone in the Firefox camp. It wasn't until recently that any acknowledgment was even made that there *was* a slight possibility of actual memory leaks, but their severity was played down and the huge memory usage still attributed to the "back button" feature.

And by "people like me", you mean people who use Firefox religiously and have been a major Mozilla supporter since Netscape spun off the browser source and actually used to be employed by Netscape before AOL came along?

It seems like this is also a typical response to any criticisms. Not only are there no "serious" memory problems, but anyone who doesn't buy into the "we need hundreds of megs for the back button caching" must be an ignorant IE-lover trying to stir shit. Simply not true.

There have been plenty of such articles and discussions right here on Slashdot, with plenty of deserved disbelief.

Re:Memory Leaks?

[McFadden]

100k? You only wish!

Most web pages seem to have images, which all together add up to more than 100k. Then there's the DOM tree, the Javascript libraries, all the script state with variables, objects, etc. There's IFRAMEs and OBJECTs.

Lots more than just the surface. I'm a web developer. Like most professionals I optimize my images for the web, and get them down to sizes which don't add up to the figure you seem to think they do. Most developers do the same. Javascript libraries are text, they're not all that big. The vast majority of sites don't even use iframes.

Typing the words "web page" into Google (first term that came into my head) brings up the following sizes for the first pages returned (44k, 52k, 13k, 17k, 76k, 12k, 37k, 52k, 21k). The definition of page size in this case is: "the sum of the file sizes for all the elements that make up a page, including the defining HTML file as well as all embedded objects (e.g., image files with GIF and JPG pictures)." Try it with as many terms as you want, I'm sure you'll get similar results. Plenty of headroom there before we even get close to 100k. Right now it looks like reality is on my side. I don't "only wish" anything, except in your imagination...

Re:Memory Leaks?

[ckaminski]

I should add though, that memory leaks are the easiest of problems to fix assuming you have good regression/unit test coverage or an application load generator. Purify + browsing google, youtube and slashdot in 500+ tabs should have fixed them up pretty nicely.

I like firefox...

[Misanthrope]

But on older systems, the sieve like memory leaks made it inoperable within a short period of time. Hopefully this will allow those of us who run legacy hardware to have a modern relatively secure web browser.

Firefox Seems To Losing Its Luster

[Anonymous Coward]

I remember the excitement when people first started using the trimmed down Firefox versions. Lean, mean, secure, and eventually the amazing array of extensions people have grown to no longer be able to do without.

Those days seem long ago now. The project needs a top to bottom rewrite to deal with orders of magnitude more demanding usage of large numbers of tabs over days or weeks at a time.

Firefox needs to:

1) Implement threading both between tab sessions and within tabs themselves

2) Bring the memory-performance balance up to par with other browsers

3) Implement some sort of standard memory/resource allocation/deallocation API for extensions so that people can bring up a standard window and see:

Tab 1: 35 megs
Tab 2: 50 megs ...
Extension 1: 500k
Extension 2: 100 megs == Zoinks!
Extension 3: 300k ...

So that memory/resource leaks can be readily identified, reported, and fixed.

The save active tabs option has helped to allow people shutdown and wipe the memory slate clean but that really is not a solution a decent piece of software should be forced to rely on.

Re:Firefox Seems To Losing Its Luster

[QuantumG]

I don't think any of us give a shit about the specifics.. let the developers sort that out, but anything that causes the browser to lock up such that you can't switch tabs needs to be fixed.

Re:Firefox Seems To Losing Its Luster

[Pulzar]

Your list of must have features are not end-user features. Why should the browser be bloated with what are debugging and profiling tools?

They *are* end-user features, though. In Windows, you can open the task manager and see how much memory each task is taking up. Would you also argue that that is a bloated debugging feature? Is 'top' a bloat? Firefox is a little OS of its own, running multiple extensions and web apps, I don't see why a feature that's standard on every OS is so non-applicable to Firefox.

Since every instance of Firefox is different because of the extensions, the only way to figure out how to keep the memory usage down is by having these memory-reporting features available. It's a necessity, as much as it is on other platforms.

Re:Firefox Seems To Losing Its Luster

[QuantumG]

You're actually complaining about 65 meg of ram? I don't know if you've been keeping track, but next year machines will come with 4 gig of ram standard.. That means you could have 63 tabs open at the same time and not even swap (assuming your 65 meg per tab thing was right, which, of course, it isn't). They're not trying to make a browser for a freakin' mobile phone here ok? Wasting time optimizing memory usage is just that, wasting time. Of course, if you really want that, go grab a copy of Opera for cell phones and use that.. or, ya know, do the memory optimization on Firefox yourself, but I'm over here not caring about the memory usage because I have a modern computer.

Re:Firefox Seems To Losing Its Luster

[calebt3]

Why not write an extension instead? That way those that would use it can and those who won't don't need to waste CPU cycles and memory on it.

Re:Firefox Seems To Losing Its Luster

[Pulzar]

Top and Task Manager are not end-user functions, to think such is to misrepresent who "end-users" are...they are not developers (if the product is to become successful with actual end-users).

Are you saying that you only use those features when developing applications?! I use them on a regular basis, to see which application is slowing me down, to kill an unresponsive task, to see if it's time to reload Firefox :)... Sure, those are "advanced" uses, but they are still end-user features. Even my dad has learned how to kill Acrobat Reader when it hangs his system, and let me tell you, he's the furthest thing from a developer.

Again, I ask you to list the functional requirements for this feature.

I think the original poster described it well. But, to summarize: I'd want to see the list of apps that Firefox is currently running and their memory usage, and to be able to kill the misbehaving ones if they won't let me shut them down themselves.

Re:Firefox Seems To Losing Its Luster

[pavera]

I've been browsing in firefox for > 3 years exclusively, all three platforms (windows, linux, OSX).

I would simply ask, what other browser has memory profiling built in? Can you open a window in IE and kill a stray activeX process or see how much memory its using?!?

Opera doesn't provide these features either.

I don't think IE is threaded by tabs, I'm sure safari isn't. I guess I don't see where firefox is so massively behind the other browsers. It doesn't use an inordinate amount of RAM, it is comparable in speed to safari, IE and opera.

Re:Firefox Seems To Losing Its Luster

[Mikey-San]

The project needs a top to bottom rewrite to deal with orders of magnitude more demanding usage of large numbers of tabs over days or weeks at a time.

Whenever I see statements like this, I ask myself, "Has this person ever done any real software development?" Rarely does a project--especially one like Firefox--need a "top to bottom rewrite", regardless of problems it's having. Even when applications make the transition from one platform to another, they almost never require a total rewrite.

Posts like yours sound really informed, what with phrases like "implement threading both between tab sessions and within tabs themselves". The reality is that in addition to not knowing that a stack of existing bugs doesn't mean "it's time for a rewrite", phrases like the one I quoted are more vague than they will appear to those who don't know better. What does "threading between tab sessions and within tabs" mean, exactly? What operations do you want to see performed in separate threads?

Firefox doesn't need a top to bottom rewrite, but I think your post does.

Re:looking forward to going back to firefox

[Seumas]

I love my firefox, but with Opera and Konq out there, the only reason I really stick to Firefox is for the extensions that I simply can't live without. I am getting so damned tired of it crashing on KDE time after time that I'm on the verge of being willing to dump it all and survive extension-free. As it is, I'll be just browsing around, reading some stuff, click a link... the page I want will start to come up... and then it'll just hang out of nowhere and never come back to life. I'll kill the process and re-launch it and it'll be fine again for a few hours. It's just so damn frustrating. Thank god for the session saver. That absolutely had to be implemented, because without it nobody would continue using firefox unless it was completely crash-free.

I do like the idea of using Konq full-time, but the extensions just aren't there. Meh.

Re:Overall, feels good and polished

[Dr_Banzai]

The Super-Autocomplete dropdown is disturbing. I'll be typing something into the URL bar trying to show my mom a web site and I'll see a few porn site entries flash by in large type and with kinky icons. The older list was much more discreet.

It should be fast

[T-Bone-T]

No matter what you want doing -- opening a new tab, moving tabs, opening up Find, zooming in and out of the page, bookmarking -- it all happens swiftly and smoothly. Those don't strike me as particularly hard things to process. Browsers have been doing most of those things quite well for a long time on much weaker hardware. If the browser bogs down adding a bookmark, it has serious problems.

Re:on leaking

[karlto]

'leaks like a sieve' has been a figure of speech for quite some time now, I don't think you can blame its inaccuracy on the author of the article

Modern attitude to bugs

[mi]

What surprises me about the Firefox 3.0 beta is how many memory leaks that Mozilla have fixed. Complaints of memory leaks with Firefox 2.0 were met with an attitude of "Leaks? What leaks?"

This is really the worst part of modern software-development practices. When users complain about bugs, they are met with hostile demands to explain exactly, how to reproduce the bug, and the complainer is always presumed to be doing something wrong. Those, who aren't willing to put up with the hostility are not even deemed worthy of being a user — if you had a bug, you should've reported it!

But when a new release has (some of) the bugs fixed, the fixes are touted as a major leap forward. We are supposed to love the new version for all the fixes it includes — and ignore all the bugs, that the next version will be addressing...

Re:Modern attitude to bugs

[mi]

Obviously, you aren't a developer. If you were, you'd know what they are dealing with.

Is it really "obvious"? Do me a favor, look me up on the Internet :) Start with the hostname part of my e-mail address.

Seriously, if you can make a mistake like this, it may explain some of the problems you are having supporting your customers...

Nowadays, I answer the phone politely but tersely, and I don't really bother to hide the fact that I have better things to do with my time. [emphasis mine -mi]

Thank you for illustrating my point.

Ack!

[Hangly Man]

And since memory is so cheap these days and everyone has a ton of it, what's the big deal about half a gig dedicated to the browser anyway?

Maybe if you're a web developer. My whole OS doesn't use half a gig of memory!

Re:Ack!

[heinousjay]

I would hope not. The OS on its own isn't really doing any useful or interesting work.

ACID2 == Microsoft Mentality == Evil

[knorthern knight]

ACID2 fits in perfectly with the Microsoft Mindset. Remember how MS screwed other browsers. They...

1) custom coded their HTML generators (e.g. Frontpage) to generate badly broken webpages, which any sane browser (Netscape, Mozilla, Firefox, Opera, Konqueror, etc) would have problems with

2) custom coded IE to handled the badly broken webpages produced by Frontpage, etc.

The net result was a World Wide Web full of pages that are "best viewed with Internet Explorer". Embracing broken "MS Extensions" is wrong. Yet the people behind ACID2 seem to think that it's a good idea that a web browser should take a badly broken webpage and guess at what the "intent" of the webpage is. What's next? A C compiler that tries to guess what you intended your program to do, rather than returning a compiler error when it encounters broken C code? The solution to broken webpages should be to fix the broken webpages.

Re:ACID2 == Microsoft Mentality == Evil

[soliptic]

Embracing broken "MS Extensions" is wrong. Yet the people behind ACID2 seem to think that it's a good idea that a web browser should take a badly broken webpage and guess at what the "intent" of the webpage is. Why on earth is this modded insightful, it's hogwash. The ACID2 test is not about browsers guessing what the "intent" of the page is, it's about browsers failing in the way the standards specify.

NB, I'm rather sceptical of the ACID2 test, for the reasons perfectly expressed in this comment, but your comment is nonsense.

"300 leaks fixed" != "Leaks like sieve"

[bcwright]

The number of leaks that exist in an application has little if any relationship to how much an application leaks memory. A single bad leak that happens often can cause enormous memory consumption, but even a large number of small leaks no one of which happens very often may not appear to leak much at all. Statements like this make me wonder if their author has ever written any nontrivial code at all.

I'm not at all saying that the Mozilla code isn't a memory hog (it's well-known that it is), nor that it doesn't exhibit the symptoms of memory leaks, which is also well-known, although as others have pointed out the issues are complex and often Mozilla gets the blame for leaks that are actually caused by third-party extensions. What I am saying is that you can't just simply count up the number of "memory leak bugs" and say whether an application leaks "a lot of memory" - sometimes the two are correlated, but by no means always.

Sheesh.

Memory Leaks - Plugins

[Mike_K]

I am a rather heavy browser user. I usually have in excess of 50 tabs open. I usually keep going back to 3 or 4 of them, continuously open new tabs and close them, and occasionally go through the ones I have open and read/close them. I normally hibernate my computer (I run XP) and shut it down only occasionally. Currently my laptop has idle time of 107 hrs and Firefox has CPU time of 2hrs. I think I last restarted Firefox to install an update.

I used to have really serious memory problems with Firefox. My memory usage would skyrocket very quickly, and I'd have to close it and reopen. This stopped a while ago when I installed FlashBlock. I rarely view flash anymore, and my memory footprint is rather stable. Right now I have VM Size of 403M - not small, but I have 4 windows and 97 tabs open. Have fairly few add-ons installed: DownloadHelper, FlashBlock, IETab, TabMixPlus and TalkBack.

I don't believe that memory leaks on Firefox are a problem, at least not on Windows. I think it is the plug-ins that are causing the problems.

Cheers,

m