Slashdot Mirror
Firefox 3 Performance Gets a Boost
jason writes "Mozilla has been working hard at making Firefox 3 faster than its predecessor, and it looks like they might be succeeding. They've recently added some significant JavaScript performance improvements that beat out all of the competition, including Opera 9.5 Beta. And it comes out to be about ten times faster than Internet Explorer 7! Things are really starting to fall into place for Firefox 3 Beta 4 which should be available in the next week or two."
Flash has more and more accessibility support, but PDF is the Page Description Format. It's meant for print output and says nothing about the meaning of the contents of the document, just how they are supposed to look on the screen and on the page.
The good thing about tag-based formats like HTML is that--provided someone's following the standard--they can be fairly easily parsed regardless of the output format. With XHTML, you can read stuff on your screen, the blind can use screen readers, and web developers can easily extract and transform elements from a given document things are good as they are.
Finally, why do you think PDF = lean and mean? Acrobat proves that a PDF reader can get hideously bloated.
Safari 3.1 is supposed to be really fast as well. How do they stack up?
The benchmark used in this article is a JavaScript benchmark, but PGO was enabled for most components of Firefox, not just the JavaScript engine. And even if only the JavaScript engine improved in speed, you'd see a speed boost despite having JavaScript disabled in web pages, since parts of Firefox itself are implemented in JavaScript.
The shareholder is always right.
AntiFA: An abbreviation for Anti First Amendment.
I have nightmare's about JavaScript being the one language to rule them all... please, let's have no such talk.
It's been my experience that the extensions and multiple open tabs cause bloat, not Firefox itself.
OS X proves that a PDF reader can be as fast as HTML. Faster in some cases -- no need to lay out and render large tables, complex CSS, etc.
Do you even lift?
These aren't the 'roids you're looking for.
I disagree that it looks bad on os x (using it right now), but it has already been addressed. If you have ff3b3, you can download the os x theme.
I have. Still ass. doesn't go lighter when it's backgrounded, stays the same dark grey as if it were foregrounded.
Open-Source seems good for getting a job 90% finished and completely ignoring the 10% polish required to make it an app of the same quality as closed-source
and does a single tcp socket's stalling not cause the whole damn thing to seize up?
Sorry, copying the bad eyecandy from Microsoft doesn't classify as an improvement in my opinion.
The learn-as-you-go menu behaviour which they copied from Windows didn't work well in Windows either. The main problem is that it causes inconsistent behaviour. Repeating something doesn't necessarily give you the same menu items. It's good for newbies who read every single line before choosing one of them, but it's very bad for people who memorize what they do so they can repeat it quickly without even looking.
Mark this as off-topic if you like. I'm responding partially to the parent comment, but mostly to its score and reason.
This is a discussion board. How can you mark someone's comment as redundant? Is this an attempt to invalidate their statement? Don't blame them when it's actually a limitation of the forum system. There is no simple way to increment an "I agree" or "I have the same problem" counter, there has to be a new comment for each person who agrees. There is no way of adding weight to a comment except by increasing child nodes, or adding as many individual argument nodes that are similar. Yes, there is already one branch in this thread that talks about the memory issues, but relax not everyone perfectly gets all their statements in exactly the right location in the discussion tree. Judge it simply on what it says, not the comments location.
For what it's worth, I agree. I also have problems with memory bloating with FF. I don't really care if they are memory leaks, or memory fragments, it's still a problem that I would like to see fixed. Unfortunately I cannot fix the problem, so I will patiently wait for the next great release of FF. I have no solution, but this is my informal bug report.
Increase in speed on JavaScript will be great. There are many times when my FF instance gets temporarily grayed out when it loads a page with lots of JavaScript. This is the window manager thinking that FF is locked up and not responding.
Gah! I don't want software that fucking "learns"! I don't want software that tries to think for me. I want software that just fucking works in the first place!
It's the KISS principle. I'd rather have stupid software that works in a clear manner than all this crap that tries to figure out what I maybe might be wanting.
The cake is a pie
>I don't want software that tries to think for me.
In which case you don't want the browser to autocomplete the URL for you at all, and the fact that it finds seemingly irrelevant matches shouldn't matter.
Threads add a whole new dimension of complexity to the engine. The "right" way to do it may not even exist.
Extensions definitely should not have direct access to the threads. It would be an absolutely terrible idea. In fact, extensions shouldn't even know that there's multithreading going on behind the scenes. At best, extensions would be able to indirectly spawn threads and manipulate the spawned threads in a roundabout manner through that context using a thread-safe API.
"If a nation expects to be ignorant and free in a state of civilization, it expects what never was and never will be."
Just because I have a couple gigs of memory doesn't mean I want Firefox to consume it all. I run more than just a web browser. I don't want a lot of caching anyway, since most pages I hit are dynamic and I don't use back very frequently. I don't want a program to look and see I have a couple gigs of memory and assume it can use it all.
-Aaron
This post is encrypted twice with ROT-13. Documenting or attempting to crack this encryption is illegal.
Or does a single tab still cause the entire browser to freeze up?
___
If you think big enough, you'll never have to do it.
Agree about ray tracing, but crypto is very relevant. If you aren't using javascript md5 in your login page, then your password is wizzing around the internet in plain text - not a good idea.
Only if you don't use https, or NTLM, or Kerberos (all browser supported mechanism that don't require javascript). Depending on javascript for encryption is silly.
MD5 is extremely broken for passing passwords about.
Mod me down and I will become more powerful than you can possibly imagine!
I've seen similar complaints, but have never been able to reproduce any such problem. I can open and close tabs all day, and Firefox does not suck up all my memory, or even a significant portion of it. I can't even run it long enough to come close to sucking up all of my memory (it would take several weeks of use every day without ever closing it). Could you explain how the rest of us could see the problem? If you do, we could report it and the problem could be fixed.
What a fool believes, he sees, no wise man has the power to reason away.
Exactly. Dumb, predictable software can be used as a tool, since it does what one tells it and nothing more. "Smart" software is unpredictable, so rather than telling it what to do and moving on, one has to tell it something, then wait to see what it did. Only when software is as smart as a human (or more) will it be as useful as dumb, predictable software.
I wouldn't say that, we have Opera,konqueror,galleon,Epiphany,links,etc...IE, if anything we have the most competition.
I do agree that firefox sucks on linux tho, although i have to give FF3 some credit for improving this.
IranAir Flight 655 never forget!
I agree with you totally. I've been toying around with the Firefox 3 Betas for a couple of weeks and I think the awesomebar is the best new feature. It's not broken -- it's different. Once you get used to it, finding pages you frequently visit becomes much easier.
Say you visited the Wikipedia page on the Tunguska event a couple of weeks. If you want to revisit the page, all you do is start typing the first few letters of "Tunguska" and the page comes to the top of the list. With the old type of address bar, you'll have to type the whole Wikipedia URL or search your browser history separately. This speedup is well worth the relatively shallow learning curve.
I find it pretty stupid to compare this feature with Windows' "adaptive menu" feature. There's only a superficial resemblance. Remember, the traditional address bar still "learns" in the way you hate by ordering URLs by the frequency with which you visit them. What is it with the Slashdot crowd and being insanely conservative about their software?
As a programmer who uses a high-level, garbage-collected language with an optimizing native compiler, I would say "the memory leak" and "the crash problem" and "the security problem" are all symptoms of using C/C++ to write a gigantic network-facing program with an embedded interpreter.
I realize that it's easier to find C/C++ compilers and C/C++ programmers, but popularity aside, this has always looked like a recipe for disaster.
So do you prevent the creation of replicators?
I would like to think one day in the future someone may invent a replicator. At that moment everyone just sells stuff is suddenly out of a job. Only the people who actually create new things/ideas/etc... will be valued.
That's the 10% that takes hard work.
"Who is the Journal of Quantum Physics going to believe?" --Stephen Hawking
So get another job, since your current one was made obsolete. Times change. Nobody needs your services any more.
> Just sayin'
Sayin' what? That we should make the world worse for anyone else by prohibiting technology use, so you can keep profiting from people depending on your obsolete manual job?
The old behavior is always better, until you start to use the new behavior.
When FF2 came out I didn't like the close buttons on the tabs, or the way that they were curved and didn't fit in with Windows' tabs, I didn't like the bland new icons, and it all seemed like a bunch of hype.
Of course now I like FF2, I like how the icons are less colorful and draw less attention, and FF3 seems like the scary new release threatening to ruin something that was perfectly good before.
Same goes for new releases of any software, from OSes to games.
// MD_Update(&m,buf,j);
There are plenty of attacks (such as CSRF or XSS) for which your AV software and a firewall are useless against. You say security isn't a concern now, until you fall victim to one of these attacks that can only be thwarted by proper security in the browser.
That's pure BS. Even high-level languages have these problems. No one has had buffer overruns on PHP served pages, but how many have been hacked? Even they crash sometimes.
... ad infinitum is still not freed. I'd almost say manual reference counting is better because at least it forces you to think about how to cleanly delete your data structures.
And finally GARBAGE COLLECTION IS NO REMEMDY FOR MEMORY ISSUES. People are claiming this all the time, so I'll repeat garbage collection is no remedy for memory issues. People claim that garbage collected languages can't leak memory, and while I admit that this is true, this guarantee is practically worthless. The definition of a memory leak is a section of memory that is no longer referenced, but has not yet been freed. Garbage collection addresses this directly by freeing memory that is no longer referenced. And you say "that's great! All our memory problems are solved!", but instead we have "object lifetime" issues. You have references to this and that in your program and the problem is you still have to carefully plan when objects should disappear, and check to ensure that you have no references to that object anywhere in any of your active data structures. If you do, everything that is referenced by that object or referenced by a reference from that object or
Furthermore, garbage collection does NOTHING about memory fragmentation. Memory fragmentation occurs because all the objects you want to allocate are different sizes. Even though when first allocated you could compact them all together, after freeing a couple and and allocating some more, you can end up a ton of memory wasted between the different objects that somehow got spaced out. What you want to allocate a 64 byte structure, too bad the largest space we have left is 62, guess we have to brk another page to put it on. Strings are particularly notorious for fragmentation (which is why Firefox already uses pooling to combat the problem). Until 'everything is a handle' replaces the modern 'everything is a pointer' in modern languages and the runtime defragments memory while the program is running (at a scandalous cost in performance), memory fragmentation will be another problem to consider.
The fact is there is no easy solution to any of these problems. You can't say `use language X, it has feature Y' and you won't have that problem. These problems are hard, but at least there is some benefit -- it justifies making that much more than minimum wage.
I've also known other people to complain about it - while others have no idea what they're talking about. You have to love sporadic issues.
I'm pretty sure it has to be a combination of minor flaws in Firefox (as not every program I run has this issue), Windows XP (with memory handling far better than previous versions, but still not exactly the gold standard of memory management), and all the myriad changes to my system's configuration over time.
It does it on both my work computer and my home computer. Then again, they're both XP Pro machines. They're very different in terms of hardware, which tentatively rules out a specific hardware config. The memory on my home machine is double that of my work machine, but the highpoint seems to be about the same - so I doubt it has much to do with total system memory. What sites I hit seems to have no relevance. The only other common factor I can think of is that I run NoScript on both - though if I remember right the problem predated my use of that (I first noticed it a while back, and had actually hoped 2 would fix it).
The main issue seems to be that a specific amount of memory is eaten up when you open a site in a tab - but closing that tab often doesn't clear up the space. I just now closed every tab except Slashdot - and it went from 157mb (when I had 14 tabs open) to a minuscule 153mb. From experience, waiting for it to dump cache is ineffective. If I close the program, the memory clears itself just fine - but only if there is no other Firefox window open. I'm guessing that multiple Firefox app windows share a footprint.
Then again, saying "all" of my memory was an exaggeration. I've rarely seen it hold on to much more than 150mb after closing all but one window, though if I go on for very long without at least shutting down firefox, that minimum can creep up - and on a few occasions really has taken more memory than I actually have on my machine (virtual cache trash time). It's also probably not a noticeable problem unless you're a heavy multitasker (in which case that footprint becomes painfully obvious).
I've noticed it doesn't happen on my fiance's Vista machine, or on any 2003 Server boxes I've run it on. It may be a problem that only occurs on XP, or it just doesn't like the way I smell.
However, if you are able to reproduce it, you'll see it happen whenever more than one tab has been opened. Opening one tab, then clearing it, seems to work - but once a second tab opens, clearing the original tab clears it's footprint, but any tab opened after that exhibits the problem.
Do not confuse "Freedom of Choice" with "Free Will".
No one will create "recordings" for sale anymore since they can be duplicated ad nauseum for squat.
There will still be a thriving market for LIVE music and plays and such things. Notice how the music industry never managed to die like it has always said it would? Started with the printing press, which is little different than me ripping a song off a CD, and now with the digitization of music and the internet. Both produced a cheap, easily duplicated copy without any harm to the original.
The only argument I can't make is for movies. They don't translate well into 'live'.
But music will simply go back to it's origins in live performances. The Grateful Dead were decades ahead of their time. Allowed free and ubiquitous copying of their performances and still made a bundle o cash.
People in cars cause accidents....accidents in cars cause people
I think instead of multi-threaded, the developers should seriously consider a multi-process model. The front end skin could broker back end processes and provide a display buffer. This would provide free page isolation. If a plug-in goes haywire, bizerk or whatever the kids says these days, the front end can just kill off the process and continue humming along.
-rd
Bah!
Those who are against replicators are just like those who sold buggy whips in the early days of automobiles, or those silly fools who took on pirates with their "intellectual property"...
Those who live by the sword, get shot by those who live by the gun...
However, I think the problem is that the Firefox memory footprint usually remains around its peak once reached, even when all but the final tab has been closed. I imagine this is due to heap fragmentation and if so, a lot of memory is wasted. Whatever the reason, it prevents the memory from being used by other applications until Firefox is closed.
Allocated RAM is no use if it's not being *used*.
Would it be fair to say the following actually occured to obtain these javascript performance improvements:
1. instrumented firefox (PGO technology)
2. ran the stinking benchmark with the instrumented code
3. used the feedback from the benchmark to automatically compile an optimized version of firefox optimized specifically for the benchmark.
4. Publish results of said benchmark for all to oooh and awwww over.
Isn't this as pathetic and useless as vendors manually tweaking their 3D drivers to artifically raise performance figures displayed in 3dmark? Did I totally misread TFA?
The fact that Firefox, being an XUL application, has significant chunks of core code written in JavaScript, is probably the main reason why this matters very much. Even if you never visit a site in the wild that uses JavaScript or run with JavaScript totally disabled, you should still see some general improvements in Firefox's performance.
Qu'on me donne six lignes écrites de la main du plus honnête homme, j'y trouverai de quoi le faire pendre.
I know, it really sucks that people don't write books anymore since the printing press came out. The whole industry collapsed, you probably haven't even heard of these jobs:
* Copyists, who dealt with basic production and correspondence
* Calligraphers, who dealt in fine book production
* Correctors, who collated and compared a finished book with the manuscript from which it had been produced
* Rubricators, who painted in the red letters
* Illuminators, who painted illustrations
Ohhhh wait, people still write books and the industry didn't collapse. It just changed. I'm sure in 50years we'll be saying 'wtf was a publisher again?'. And nothing of value will be lost. Artists have the HUGE opportunity of being able to cut out the middle men (there are lots of them) with current technology. With less hands in their pockets they will make big money from live shows and bigger profit from merchandise as well as profits from ad supported downloads and site page views. Artists will NOT starve, i don't see how cutting away the massive corporations which artists are carrying on their back atm will hurt the artists.
Not allowing the extensions to be multithreaded means that you've lost most of the benefit because your extension threads soon become single points of contention causing the multiple browser threads to stall.
There are 4 boxes to use in the defense of liberty: soap, ballot, jury, ammo. Use in that order. Starting now.
I'm not trying to be partial, I did search Opera's Web site and the most recent version I could find for download is 9.50 beta. Unfortunately, Opera 9.50beta failed completely. It would not even finish the test (hung on crypto:aes). That is why I used vesion 9.26 (it being the most recent aside from 9.5 on their download page). I mentioned this to a friend, however, and he said it failed for him the first time, but worked a subsequent attempt. So I just tried it three more times, and one of them it actually completed. It resulted in the following:
I'm happy to try a more recent version on this same hardware if you can point me to one and if it actually is functional enough to run. Likewise if anyone can point to another browser they would like me to reference for them.
P.S. if you're involved in Opera development, the fact that selecting the "Opera" menu and "About Opera" loads a Web page over what you're doing, in my case canceling the test and making me start over, is really annoying. Can't it at least load a new tab?
Oh, and just for fun I also ran a few other browsers:
That makes the full list for OS X:
I'm sure I missed a few and there are probably newer betas of some of those, but they were not easily found on their Web sites.
It does. Your "just fucking works" is slightly different from my "just fucking works", so it learns how we each work and adapts accordingly. It not only learns your habits in entering URLs in the address bar, but it also learns from your browsing history and bookmark use. When I type "sl" lo and behold, Slashdot is the first entry. When I type "gl" the first entry is for my Globe And Mail portfolio listing, when I type "bm" my online banking page comes up #1. Now, if I wanted to view other pages that I select often I can scroll further down, but in general all I need is two letters, down, enter and I'm at the page I want based on my own browsing habits. Now, your online banking might start with "sc" or "ba" or any other combination so you wouldn't want "bm" associated with that. Maybe "bm" links to your social networking site or something else you want. Why bother with one-size-fits-all when you can have custom tailored for the same price?
BD Phone Home!
Shameless plug. Like you weren't expecting it.
I beg to differ -- Webkit is actually pretty damned good, in terms of speed, stability, and compliance. About the only thing it lacks is universal support from webpages, but then, that used to be a problem for Firefox -- there's not a whole lot a browser can (or should!) do about sites that don't follow the standards.
I'm going to say IE. Whatever replaces HTML must work in IE, and it must do so without any fuss.
But that's only because I'm assuming we'd replace HTML with something like XHTML, which is (inexplicably) broken in IE.
PDF is no more or less "fully graphical" than HTML. In fact, if I remember right, PDF is based on PostScript, which is a Turing-complete language designed to format documents for printing -- sounds pretty textual to me.
Flash might be, but do you really want the Internet to be based around keyframe animations?
The difference is, HTML is designed for the Web.
PDF is designed for print. I'd much rather have a website which I can resize to any window I want, and let the text flow to fit, than a PDF document which has pages that are certain proportions, exactly, leaving huge margins and gaps between pages for no reason. PDFs certainly have their place, but the Web ain't it.
And I only say that because I don't know enough about the other capabilities of PDF to know if it can quite replace the Web as we have it -- plugins, javascript, video, etc.
Now consider Flash. Adobe certainly seems to want this to happen. AIR embeds a SQLite engine, a Webkit engine, and probably some other things as well, basically allowing you to develop a desktop app in HTML and Flash -- but since it's Webkit everywhere, you don't have to worry about browser compatibility. And Flash itself keeps getting more and more capabilities, though it still sucks for video, and still isn't hardware-accelerated at all in the browser window.
But it's really frightening that you want to replace HTML, an open standard that works just about everywhere, with Flash, a wholly proprietary system that is only relevant to this discussion because Adobe has seen fit to port it to enough platforms. I'm still waiting for my 64-bit Linux Flash, and I can't do a damned thing about it, other than contribute to reverse-engineering projects like Gnash.
And yes, I know you can get the SWF spec. You can get it under a license which forbids you from developing a player. Woo hoo.
Stop right there.
You want us to stop working on improving HTML engines, and start working on adding features that don't currently exist to PDF or Flash? And then write a browser around them?
Think about that for a minute. Do you really expect it to be easier that way?
HTML is easy to parse. It's even easy to render. It's broken HTML that's hard.
And XHTML would be even easier -- but again, there's that Internet Explorer problem. That, and most websites aren't ready, either.
Now, if your suggestion was to start over completely, I'd be all for it. There are many things I wish had been done differently. But that would take an order of magnitude more time and effort than your crazy-assed PDF and SWF ideas.
Don't thank God, thank a doctor!
How can this be insightful? It's stupid and insane, and makes no sense.
I don't want programs that tries to think for me, but I want it to present options if possible. Autocomplete a URL is like tab-completion in the shell. This is useful and consistent.
Trying to present a selection of things in a mozilla-programmer-logic order will almost never be what the people want.
Stop telling people that their argument is invalid by being a straw man, damn it!
In IE7 I really don't see the importance of the about:tabs or about:blank homepage. It's not unsettable as far as I can tell, the text is always there. Sure, it's selected, but sometimes I click in the address bar and have to remove the text. A nitpick, surely, but annoying. Start with a blank screen and a blank address bar. What can be so hard?
No...I want autocomplete. What I don't want is autocomplete based on arcane and mysterious rules that I do not understand. I want autocomplete based on simple and clearcut rules, like "suggest the most recently used URL that matches". In my experience, software that tries to "learn" is harder to use because you end up trying to figure out how to get the damn thing to "learn" what you want it to learn.
The cake is a pie