The Firefox 3.5 fiasco

(updated: replaced 'trashing' with 'thrashing' as indeed, I meant 'disk thrashing').

As a Firefox user, I was delighted when Mozilla released Firefox v3.5. It was advertised as a new milestone in browsing, with more standards being supported, new engines for javascript and web content rendering and the intarweb would appear to me faster than ever before. As a person who has a bit of a blind spot for marketing and everything related I was a tad skeptical, but I thought "What the heck!".

So I went ahead and downloaded the installer at release day and after fighting with the usual plug-in upgrade mess, I was able to run the browser for the first time and lo' and behold, the web felt like I was back in 1994, when no-one but the Real Geeks had web sites and everything was lighting fast. Live was good.

The next day, with a fresh cup of coffee in my hand I started my beloved Firefox 3.5 browser on my freshly booted system. I was expecting to see the browser dialog within seconds to re-experience the web at light speed, but nothing happened. Well, something did happen, my PC's hard-disk was busy like I was running three virus-scan sessions at the same time. After 35 seconds or so, it finally managed to find all the bits and pieces it apparently needed and showed me the familiar face of the Firefox browser dialog and I was on my way to the outside world!

Suddenly, a small, screechy voice in the back of my head tried to make a point. That voice, which sometimes cries through a developer's head when s/he writes a piece of code which isn't in the format the voice owner likes, at which point it desperately tries to convince you to do something else instead by giving unwanted advice like "Wouldn't it be better if... " and other lovely comments no-one wants to hear, that voice made a remark in it's usual dull way about those 35-something seconds before the browser really started. As with similar occasions, I didn't pay much attention to it. Every Firefox instance I started was lighting fast, and showed up 2 seconds or even faster, it must have been something else which had caused the delay at startup. I know, it didn't sound very convincing the first time either.

That afternoon, I had no browser window open and started a new one. Again I was rewarded with a long pause, disk crunching and a blank screen, until 30-35 seconds had passed and Firefox 3.5 was awake and ready for duty. "Hmmm..." I thought. Voicy in the back of my head was awake again too, with random babbling about "Told you so", "I'm not gonna repeat myself" and similar wisdom, the usual. Could it be windows or some service caused all this disk thrashing and the delay? I more and more got the feeling Voicy was right (I hate that feeling) and there was something fishy about all this.

I didn't want to wait 35 seconds every time I started a browser, so I wondered what to do. Then I realized I was an end-user of this application, this browser. And what do end-users do? That's right, they go to the support offering of the vendor. Mozilla has a nice forum system so I searched it a bit to see if fellow Firefox users had similar delays. Well... you could say... yes indeed. And not only delays of 30-40 seconds but some had to wait minutes or even worse: after starting Firefox, it went into a coma and never truly woke up. Mozilla also found out that more and more people had the same problem and added a sticky thread to their forum. You can reach it here.

That forum thread revealed what the true cause was of this disk thrashing and delay at startup. I have to warn you though. If you're a developer, your software engineering fire will die a little when you read the true cause and from then on you will have to fight off thoughts of giving up development altogether and apply for a job in marketing or HR. So what was it, what's the cause of this slowness? It's NSS. What? The Network Security System. It turns out that NSS needs to do all kinds of encryption and other security related tasks (which seems kind of logical), and for that it needs random numbers. Sounds reasonable, right? Well, it kind of does.

True random numbers are hard to produce, because in a computer system, nothing is really random, it all is a result of some action which was a result of some action etc. etc. The clever boys and girls of the NSS team had to crack this problem: how to get 'true' random numbers which are as random as possible? Instead of using the randomization functionality of the underlying operating system (which has this feature build-in as every TCP stack for example needs it), they did what Mozilla in general always does: they re-invented the wheel. Nothing against re-inventing stuff, don't get me wrong, not every wheel is as equal as the other one, and you can never have enough good, re-invented, shiny wheels. Though, the downside of re-inventing wheels is that along the way you can't make mistakes, it has to be better than the previous invented wheels. No-one wants to use your square new wheel for example.

To solve the problem of the randomization, the NSS team came up with something clever, something so great, that no-one else had ever thought of that before: they decided to read the files in all possible temp folders on disk with multiple threads so these files can be used as seeds for the randomization. Brilliant. Temp folders! Why hasn't anyone else thought of using a disk-based resource for random number generation! I mean, these folders change every couple of milliseconds, have immediate access, no latency to read their contents and are never filled to the brim with useless cruft!

That is, if you're on the NSS team. In the outside world, things are a tad different. You see, Firefox v3.5 reads the Internet Explorer Cache and the central Windows temp folder in your user profile, through its NSS subsystem. Not only is it, in my humble opinion, not done to read another application's caches or temp folders, it's also amazingly ignorant towards the real bottlenecks of our modern computers: hard-drives. If you're using a virus-scanner which is set to paranoia mode, this whole temp folder traversal by NSS will be even slower because every file accessed will be scanned by the virus scanner. Over and over and over again. And what happens if the user doesn't do anything else but browse with Firefox, so these temp folders will not change (or are empty)? Isn't using file reading the worst way to obtain a seed for randomization?

I used sysinternals' Filemon tool to check which folders and files Firefox was reading and along the way I also saw they read all fonts up front. All of them. That too seems rather odd for a browser who claims to be the fastest browser. How many fonts do you need on a random (pun intended) webpage? Besides the default ones and a few common ones? 2, 3? Would it hurt anyone if these are read 'on the fly'? Not compared to the delay in startup time for a browser dialog when you have many fonts installed.

NSS is open source, but it's not something you can fix yourself, unless you compile the browser as well. The problem is that NSS is a security component and therefore needs to be signed by Mozilla to be used in Firefox. This means that recompiling the NSS dlls won't work, Firefox won't accept them (which is logical, it's a vital part of the security system, heck it is the security system!). Though, why should I even bother? It's 2009, for crying out loud. After 15 years of web-browser development, the human race should have produced a web-browser by now which is worth using, without silly startup delays which last minutes or even longer. After all, in this case, I'm an end-user.

Mozilla on their forum says 'a' developer is working on a fix and they 'hope' that this developer is able to fix it. That's not sounding very promising to me. This is a top priority issue, Mozilla, unless you want droves of people drop your browser for the competition. There's already a fix available, Firefox v3.0 didn't do this disk thrashing and is able to communicate security over the internet, at least that's what you always told your users. In other words: the NSS version in Firefox 3.0 was capable of creating random numbers and doing encryption without the necessity of reading a competing browser's disk cache nor the OS' temp folders. In case you wonder, Mozilla, no, I'm not going to advice friends and family members to use Firefox 3.5 anymore till this is fixed. Not that you nor your droves of developers lose any sleep over that, at least I hope not, but with me I'm pretty sure more people will do the same: move away from Firefox or revert back to an older version and wait with advice to friends and family about Firefox 3.5.

I'll revert back to Firefox 3.0 till this is fixed, or move to another browser (although I find Chrome a bit too much Google in one package). If you're planning to upgrade to Firefox 3.5, be aware of the issue I described above and do realize that it's not something you can learn to live with, as the delay will occur randomly (pun intended) during the day: sometimes starting a browser is fast, however an hour later it can take again 30-40 seconds or longer.

Yes Voicy, I'll listen to you more. At least more often.

138 Comments

  • I hate to sound like an MS fanboy, but if you haven't used it yet, IE8 is a great offering. It's much faster than FF3 (not sure how it compares to FF3.5), has some nice additions compared to IE7, and the developer tools are great! They're *almost* as good as FireBug, but are a lot easier to use.

  • No such problems on the Mac version. There was some weirdness caused by the Google toolbar, but uninstalling and reinstalling that cleared it up.

    Weird though that they'd end up taking a step back like that.

  • Wow. Just, wow. And you know what? I just started Firefox and watched it in the Resource Monitor. Sure enough, it appeared to touch every single damn file in the fonts folder.

    This is a completely amateur mistake. How can we trust a security system that is built by somebody making mistakes like THIS?

    My God...

  • FireFox 3.5 seems just fine on the systems I use(WinXP and MacOSX).

    Btw, why do you have that awful Windows font on your page? Did you know that for those(like me right now) that don't have that Cambria or what-ever it is called, the page is unreadably compressed.

  • @mac user: it's a normal font? I wasn't aware that this page uses a stylesheet with a font that wasn't available anywhere else. (I based the style on an existing style of this blogengine here at asp.net). I'll check if I can change it.

    FF 3.5 seems to suffer from the slowdown delay only on windows indeed. Which is makes it even more weird why on windows NSS uses these odd file-system based randomization. Anyway, I installed v3.0 over 3.5 and everything is back to normal

  • Must only affect certain users. I have no such issues on any of my PCs. I use it on 3 different computers and it never delays like that. Then again, I never use IE so nothing to scan there and I keep my temp files pretty cleaned out. I also used the beta 3.1-3.5 on all 3. I like my ff3.5

  • no problems on Ubuntu on 9.04

  • Chrome, baby. It's lighting quick.

  • well, on Linux you don't run an anti-virus, but anyway the OS gives you random .. so Fx 3.5 is just fine.

    Now if only the graphic drivers were better ... :)

  • Give Opera a spin, it's by far the browser that uses the least resources out of the bunch and also the one that comes with the most features by default. The thing even has an adblocker and a web server. I started using it a few years ago and I really can't go back to firefox anymore.

  • Through a pretty amazing co-incidence, I too was running sysinternals' Filemon tool (for the first time) LAST NIGHT, and saw this exact same thing happening: "why the hell is FF accessing the IE temp folder? Virus? Snoopy WTF? This is not right"...so here it is on your page. Thank you. Now, Mozilla people - are you friggen serious? Big bloody letters on the wall: MISTAKE!
    Take the startup lead from Chrome...

  • @mark: Even though IE 8 start up rather fast and appears pretty snippy to use, try and run the tracemonkey or any other JavaScript or DOM selector benchmark. It is many many times slower than Firefox or Chrome. I know. I get to develop cross-browser web apps where 80 percent of my effort goes into performance tweaking and css hacking for the IE family.

  • Hey... Stop bitching and get in there and contribute something to the project yourselves... You don't like what you got for free... Tough! Move on or contribute yourselves...

  • Actually, this is how I feel...
    And before i rant here, I use firefox 3.5 and have no real problem with it.

    But my real love came with firefox 0.x something i think 0.6 or so, I dont really recall.

    I was using IE back then and hey, firefox beat IE easily. But after so many releases, they added a lot of things which I simply didnt need at all. I dont really care about all the stuff that is added.

    I dont want a million options. I dont want an operating system becoming my browser .... where are the good old days.

    What I want of a browser is that I can browser sites quickly, efficiently, at best without any stupid popups molesting me, or idiotic flash advertisement jumping at me. The new button to create a new tab is something I dont need at all, and in fact i hate that new button. I dont want it, I dont need it, and we can disable it in user*css something.

    But still - why do I *want* to disable all what I dont want. The future releases wont really focus on "powerusers" who just want to browse in order to gain information.

    I would really like a super fast browser, which I can customize a LOT. Something that starts lean, and where I slowly add components, until I am happy.

    That would be a good browser.

    Of course, the Internet Explorer is dead since many years, and Opera is a good browser but it also has a few other problems (read: sucks in different ways)

    I am a lost soul of the internet.

    Firefox is NOT bad at all. But it could be A LOT better if it would stop trying to treat ALL users the same.

    PS: I dont care about security imposed on me btw, and I also hate the auto-update feature. I think it is a joke that a company wants to *think* _for_ me. I can understand that the majority of users is probably an "average joe" who will benefit from the way mozilla develops, but I DO NOT WANTED TO BE PATRONIZED BY THEM. Or anyone else.

    Can we please have a much leaner version of firefox where we add components how we wish, but without the stupid .xpi addons? Like a TOTALLY modular browser?

  • ...Just get Opera ffs...

  • Really interesting info. I always wondered what the heck Firefox was doing that it took ages to load while stress testing my processor.

    Btw, no offense, but this article is almost unreadable with the way the type is set. Try this, please:

    div.post p, #comments p {
    line-height: 1.5;
    padding: 0.5em;
    width: 70%;
    }

    Cheers!

  • Security madness has gone too far. I remember back in the days, when I was working with a network security company and everyone was instructed to set their browser to max security level. I remember watching people clicking "allow" at least two or three time for every page they loaded. Nobody would ever have noticed any malware..

  • After read all that I cleaned up the temp folders with ccleaner and ff3.5 takes about 2 sec to open.
    That seems to be the "temporary" solution: keep it clean. Obviously it needs a better solution by the NSS.

  • Seems fine to me, I have shed loads of extensions installed (and "nightly tester tools") firefox opens around 150 tabs at start up, so it's a while before it's really responsive anyway.

    @A Mac user is right though, the font is horribly compressed.

  • It was slow for me at first, but when i uninstalled some uneeded add-on extensions, it speeded up considerably for me.

  • people do stupid things like this and who gets the blame ? Windows and then some times microsoft.

    Firefox should sort this out it is my fav browser these days and I dont want to change that. May be IE might become my default and fav gain.

  • I use FF 3.5 and I don't have this issue. However, I forcibly clean all temp files when FF exits. As well, IE does not get executed on my systems. I have waited a few seconds, (5-6), before FF starts, but nowhere near the 30+ seconds mentioned in the article. Try using ccleaner or something similar. You should in any case for security reasons :)

  • "the human race should have produced a web-browser by now which is worth using, without silly startup delays which last minutes or even longer."

    They have, it's been around since the '90s. It's called Opera.

  • I had a different problem with Firefox 3.5. I left my WinXP computer on last night with an instance of Firefox open to iGoogle. When I got into work today I had a system warning me that virtual memory was dangerously low. I took a look at the Task Manger/Performance Monitor and saw the Page File was at 1.8 GB and CPU was running at 50-70%. i shut down Firefox and it got better, but I rebooted anyway. Is that a symptom of a memory leak?

  • P.S. Were any of the above commentors involved in helping out with the betas that led up to the release of 3.5? And what steps have you done to help remedy any problems that you have found apart from bitching on a public forum.

  • everything works just fine on my end no start up issues whatsoever , also as someone pointed out, you would think about little things as using the proper fonts on your blog, this is an unreadable mess

  • @Mark: modern web standards support? Webkit, Gecko and Opera are the only contenders right now.

  • P
    {
    font-size: 10pt;
    line-height: 10pt;
    }

    WHY WOULD YOU DO THIS? Ow, my poor eyes...

  • @mark, unless you know, you want a browser that renders things to the spec of the w3c. Granted IE8 is closer, but Safari, Chrome, and Firefox are much better at rendering to spec.

  • "the NSS version in Firefox 3.0 was capable of creating random numbers and doing encryption without the necessity of reading a competing browser's disk cache nor the OS' temp folders."

    Actually, no. 3.0 and 3.5 both read temp files for entropy, but read different temp dirs and in a different way. From reading the bug in question, their automated tests didn't catch the slowdown because they didn't have an IE cache filled up.

  • Nooooooo! What a horrible idea. Check Game Gems I or II for better ways to generate true random numbers. I don't know why they don't just build in a true random number generator into the CPU circuitry.

  • Or... you could clean out your temp stuff. I've noticed no slowness whatsoever since upgrading to 3.5 on release day, but I also use CCleaner to empty out my temporary files at least once a day.

  • That is retarded, it's not even random. Whoever did this should be humiliated and expelled from the software industry.

  • If you're uising a viurs scanner you don't get to complain about speed, ever.

  • isn't this the b1 version of 3.5? so why are people acting like thins is unexpected?

  • isn't this the b1 version of 3.5? so why are people acting like this is unexpected?

  • I use XP and Win 7 on a late 3.2ghz Pentium 4 and a 1.8ghz Core duo respectively and I haven't had nearly the problems you have. It is a little slower and does read the disk heavily but it starts up in just a couple seconds.

  • I'm on Vista and get maybe a 1.5 second delay between starting FF3.5 and actually being able to use it.

    In fact it's almost the opposite case of 3.0 vs 3.5 for me. 3.0 on my other computer is a turtle. Takes 20+ seconds to be usable.

  • Sorry, but after trying Minefield in the past I wondered how FF 3.5 could suddenly change. So after reading your article I downloaded 3.5 installed it and it opened faster then 2 sec. (2 sec because I have Vista and I have to say OK to Java.)
    You are an IT guy but something is seriously wrong with your install, system or just maybe you have a MS gold nose ring.

  • LOL @ using Windows. What a joke. And someone mentioned IE8?! Developer tools are 'great'!? Oh man, what a barrel of laughs the internets are today.

  • This is a "fiasco"? A browser taking half a minute to start?

    As for the issue: using temp files to initialize a PRNG is stupid, but if you just leave the browser open it's a non-issue. My cheapest, least powerful machine (A eee 900) has a gig of ram. I leave firefox open on it. Are you so starved for ram that you have to close firefox when you're not using it?

  • I use 3.5 on XP pro and have no problems whatsoever. It takes a maximum of 7-8 seconds when run after a restart but is almost instant after that. I even tried using IE8 but found FF faster.
    I use Excel, VLC, Utorrent and FF for 90% of my time on the PC. Just average stuff. Probably similar usage to alot of people who are having problems - so, I dunno ha!

  • as a FF user, the IE cache would be pretty static...

  • Boo hoo early adopter syndrome. Mozilla doesn't have millions to spend on QA. I'm sure the problem will be fixed soon.

    IE Developer Tools are awful. I dread any time there's a bug in my code that is only in IE only.

  • In your css, change:

    line-height: 10pt;

    to:

    line-height: 150%;

    Much better :)

  • I noticed this in earlier versions (though i don't know if its the security thing), it takes forever to start, then there's the inevitable updates to do, and then finally you can navigate to a website. And on top of that it still eats up huge amounts of memory. Personally i don't see what all the fuss is about, it's just a web browser and there's others which do the job just as well without the delays.

  • When installing a Windows OS it will automatically allot a percentage of your HD to temp files and internet cache. In the past with older HDs that were tiny in size compared to todays HDs the cache wasn't that big a deal. However as disk size grew the cache got bigger and bigger. And having a large cache isn't always a good thing. I noticed about a decade ago that IE cache folder and windows temp folder could be set to a specific size and I started doing that. This does several things the most important being that it doesn't clutter up my HD. As a result of having small temp folders on my HD i have yet to experience slow start ups with FF3.5. Perhaps the developers that wrote FF3.5 also manage their file structures rather then allow Wondows to waste HD space with interenet clutter?

    just a thought.

  • been using FF 3.5 for a week and haven't had a single problem, starts up in a snap for me.

    explain.

  • I load a batch of news site files at one time and find that 3.5 won't let me look at any of them until they are all loaded.

  • Is someone going on about the security team at Mozilla? To be fair, what is the alternative for Windows in that case? Safari? Opera? For sure as hell you can't say that you have faith in the Internet Explorer team after all these years.

  • And this why I NEVER d-load the new ohh-ahh, latest and greatest upgrades until I hear that they worked the bugs out,,,,because there is ALWAYS something up with them.

  • "If you're using a virus-scanner which is set to paranoia mode, this whole temp folder traversal by NSS will be even slower because every file accessed will be scanned by the virus scanner. Over and over and over again."

    If your using a virus scanner set to paranoia mode, then you have other issues to worry about. Know what your doing and quite possibly you could get by with NO anti-virus.. or at least one with normal configuration. What kind of person in their right mind would even try to consider themselves "techy" and possibly come up with an idea to test a Paranoia config. Noobs!

  • I ran into the startup delay with a friend's computer. I didn't know what the cause was; initial FF troubleshooting of clearing the profile and starting fresh didn't work. Using FF portable didn't work either, so I also fired up filemon. My friend had millions of files in his temporary internet files (apparently the tversity gui generates an xml every time the screen is refreshed, 5 sec default and continues to do this when minimized in the systray). Wiping his temp files, which took hours, resolved the startup issue.

    I'm glad to see a root cause for the issue. I couldn't figure out WHY FF was pouring over IE's temp files on startup - it just didn't make any sense. :facepalm:

  • I use(d) FF3.5 on Win7, and the memory for this program skyrockets within instants of simply loading. It is very annoying because it slows everything down.

    One of the reasons I like IE8 is that simply closing a tab can reduce the memory due to the multi-threading.

  • Seems fine on my Linux machine...I think /tmp is in memory, not on the hard disk anyway, on my setup.

  • Frans, give the Safari 4 browser a try. It's faster than anything else I've tried.

  • Works fine on Linux.

  • >I'm not going to advice friends

    ADVISE

  • Thanks for the post. Have to say, as someone else has mentioned, the font on your blog is exceptionally hard to read, especially as you seem to have reduced the leading. I resorted to disabling stylesheets to read the page. Good luck fixing this. Toby

  • Runs fine on Ubuntu too. Fires up in a couple of seconds. I second the font criticism, I have Verdana but the lines are quite compressed. If you remove the line height from the p elememt in /DynamicStyle.aspx?SectionID=168 css file it looks fine.

    P
    {
    padding-right: 0px;
    margin-top: 0px;
    padding-left: 0px;
    font-size: 10pt;
    padding-bottom: 0px;
    color: #000000;
    line-height: 10pt; <<<<
    padding-top: 0px;
    }

  • Who said the /tmp isnt used in linux or mac to make the rnd seed? It sounds like windows has a problem with aggressive virus scanners and the tmp dir in general.

  • windows, xp & 7.. no issues at all. not a priority issue for me. speak for yourself.

  • Don't you mean "disk thrashing" with a 'th' and not "disk trashing".

  • Yet another reason to move to Chrome. (If only they'd ever get a decent ad-blocker)

  • If it's scanning temp files to use the amount of time it takes to read them as random data, that's fine. Though it should certainly limit itself to a reasonable amount of randomization and then stop. Using the file data itself, however, it utterly and completely broken.

    And, of course, Windows has 'CryptGenRandom'. And FireFox could keep its own entropy pool on disk (which you write out from time to time to ensure you never start with the same data twice).

    Sad. Truly sad.

  • I don't know how they implemented it, but in *nix they have a system wide random(ish) number generator, /dev/random and /dev/urandom. I think FF used the system random number generator in *nix rather than whatever the bastardized contraption is here, which is why they may not be experiencing this flaw. Works fine in Linux.

  • It is more likely a user problem. I have had NO problems.

  • This sounds incorrect. In fact I'm almost certain it's wrong from experience.

    When you upgraded, something changed and it causing a big loop that has to timeout in 3.5, maybe a misbehaving plugin.

    To prove it's your upgraded configuration that's causing the startup delay, try making an empty (clean) profile and temporarily try that and time it. I bet it starts in only several seconds.

    Instructions:
    support.mozilla.com/en-US/kb/Basic+troubleshooting#Make_a_new_profile

  • The squished text isn't a font issue, but it is a problem with your stylesheet.

    You've got
    p {
    font-size: 10pt;
    line-height: 10pt;
    }

    line-height ("leading" in traditional typography) should be 2-5 pts larger than the font-size, or at least 120% of the font size.

  • FF 3.5 is working perfectly on my system (windows XP). FF forever!

  • Haha brilliant one-liner there mate: "No-one wants to use your square new wheel for example."

  • running Vista basic edition here. absolutely no problems with 3.5.

  • I don't really see how one bug is a fisaco.

  • So that's the reason :). This should be listed at the daily WTF! LOL

    Anyway, I use FF3.x when developing websites as it is the *best* development browser (IMHO) and I use the Chrome Dev Channel build (currently at 3.0.192.1) which kicks all browsers asses in start up speed, stability and web load speed.

    O.. I use IE8 too to test if the website changed work in IE8 and amazingly they do :) so that's why we still need to test with IE7 and IE6 (yes IE6.. which STILL gets a way too big piece of the pie). At least we now have the rule that our applications must work but it is allowed for them to have small graphical glitches.

  • Sounds kinda weird that FF developers would make such a mistake, but I have no doubt it will get fixed soon.
    In the mean time why not try Opera?
    And as per Chrome, try Iron for a not-so-much-google experience.

  • I agree that reading hard disk is not a bright idea. But i don't see why i haven't had this issue with FF35 at all. I've been using it since the day it was released (and its RC before that) and never encountered this issue. Only time when it takes more than 10 seconds to start is when i try to start it immediately after booting - while Vista is still busy with its own start up tasks. I am on a Vista SP1 with 2GB ram. I rarely use IE and my temp folder has items that (probably) ff put there.

  • @Frans: Make even a very slight critisism of FF and you can garantee you'll get a flood of hurt comments from upset developers. It's not Microsoft, it must be perfect right? Also, font looks much better.

    @Mads: "I remember watching people clicking "allow" at least two or three time for every page they loaded" - They had UAC back then? ;)

    @Mark: (RE IE8) "The developer tools are great! They're *almost* as good as FireBug" - No... no!

    I used to be a huge fan of FF back in its lightweight days but lateley it just seems to me to have become more and more bloated and slow. If it wasn't for the mighty Firebug, I would probably exclusively use Chrome.

  • "the web felt like I was back in 1994, when no-one but the Real Geeks had web sites and everything was lighting fast."

    Lightning fast? you must come from a different 1994 then I did ....


    Also, it seems kind of logical to me but how about try cleaning those temp folders?

  • This has been happening on Vista for me for ages. FF3 took up to 4 minutes to start and FF3.5 is no better. All is good on XP though.

  • all you have to do is clean out your temp files. i had this problem too. then i used ccleaner, like some others have said.

    now ff3.5 boots just as fast as chrome did.

  • @Dan!: The point is, you shouldn't HAVE to clean out your temp files (and especially not you IE cache, which you might want for performace purposes) in order for another unrelated program to run efficiently.

  • OMG! And I was about to buy a new ssd for my EEE-PC because firefox took several minutes to start and I wondered what's being accessed by firefox on my ssd all the time...I thought my ssd is broken...now I know. Thanks for letting me know. Firefox 3.5 is really unusable on my eee-pc...my eee-pc boots fast and I want to use the webbrowser instantly after that...with firefox 3.5 it's simply impossible.

    Thanks for sharing the info...I will downgrade for my eee-pc.

  • Boy, this sounds familiar. Back in 1998, I was working at Netscape, and they had me looking for ways to speed up startup. I found some little things, but the big one was that NSS was calling popen("netstat -i") to use the packet counts as input to its seed. Now, of course, popen() does a fork(). In the Unices of the day, fork() didn't do copy-on-write, so it had to copy all the process's pages. For a large process like Communicator, this would take about two seconds.

    I proposed taking out the popen(); I was told to talk to X, on the security team. X, naturally, told me that would be insecure. When pressed, he admitted he meant "less secure"; but he still wouldn't budge. I asked why not get the packet counts directly; he said it wasn't portable--some Unices even required you to be root. (There were more viable Unices in those days; IIRC, we supported at least 5 of them.)

    I think I considered writing a popen() that did vfork() instead; but then, as so often at Netscape, I got reorged, and it wasn't my job any more.

  • No problems at all here with FF3.5, running vista, smooth and fast.

  • Crazy! I wonder how many other applications do a similar thing. It's no wonder the performance of Windows decays over time.

  • Defragment your hard drive. It will help, but it still might not be fast.

  • @John Stracke: heh, great anecdote :)

    As you see, nothing really changed: they still cram a lot of OS-specific code into NSS to be... OS independent, but at a silly way: the OS-specific code isn't calling the native OS methods to do what they want, they do it 'themselves' as that's 'better' it seems.

    Also funny that a netstat -i is used, '-i' isn't a valid argument on windows for netstat (should have been -e). ;)

  • Firefox 3.5 on my machine(Vista home basic) starts up in less than a second. I do not hear any disc trashing at all. I set it to clean all private data on exit. Would that be the reason? otherwise it all works fine here.

  • For me slow startup is not a big issue... Im starting my browser once in the morning and use it whole day..

  • Ive never had this problem. 3.5 has been the fastest to start up for me.

    Its just as easy to clean your temp folders. Run a program like CCleaner every now and then and gain performance and disk space.

  • Everything is OK on my machine too...

    Maybe Chrome faster, but I couldn't feel it.
    Speed should be tested to say definitely.

  • To everyone saying "just clean your temp files"

    Are you on crack?

    It is an amateur bug in Mozilla that shouldn't be "fixed" by every user cleaning their cache every time they want to launch the application.

    The reason why it's an amateur bug is that whoever came up with the idea didn't understand that the temp directory can contain nearly unlimited number of files. Nor did s/he understand how anti-virus applications would react to the crypto scheme.

    It's broken. It needs fixing.

  • I like Firefox 3.5 a lot, and as with you I was really happy with performance the first time I ran it. However the slow loading makes me reach for the Google Chrome icon every time my computer starts, as its pretty much instant. I do everything in the cloud and Windoze takes long enough to boot without having to wait for Firefox to do its stuff too.

  • At work I have to use IE7 with a bunch of security crap turned on - it's on an Air Force base so it's an Air Force computer. Compared to FF, IE7 in this configuration works in a very unpredictable manner:

    Middle-click on a link & it opens as a tab in a different IE window.

    Left-click on a link & it may create a new window when you simply wanted to open in the current window.

    Google up MOTW & try to understand what Microquack is describing. UGH!

    I can only hope IE8 works better, tho it will likely be years before it gets to an AF computer.

    Bob

  • Opera works just fine. If you care to take the time to learn what it can do, you won't be disappointed.
    And there are some other fine browsers out there, no question - SeaMonkey, Konqueron, and K-Meleon.
    Security? Check them out in Secunia and compare.

  • The only fiasco here is Microsoft. It's because of their lousy "random" number generators that people have to do these sort of things!

  • No problems with 3.5,Chrome is good too.I really dont like the IE7 but the version 8 is pretty good.Maybe the writers hdd is at his and ? :))) LOL

  • Black text + dark blue background = murder on my eyes.

    I hope you don't consider yourself a web developer.

  • FYI There is a very active blocker bug about this with Mozilla:
    https://bugzilla.mozilla.org/show_bug.cgi?id=501605

  • @Chris: I think you blocked some essential background picture(s) ;) The background of the text is light grey/gray. And no, I'm not a webdesigner, I write o/r mapper engine code for a living.

  • I used 3.5 RC1 and 3.5 RC2 without a problem, but the final release of 3.5 just doesn't want to work for me. I just removes firefox altogether and reinstalled an older (3.0.11) version I have, it works fine now.

  • I'm using 3.5 on Windows 7 and it works perfectly and fast, 2-3 seconds.

  • Works for me, no issues on XP, Vista, and Windows 7 - tried it out on all of them.

  • Well, that explains why firing up firefox on my desktop was hanging so badly tonight.

  • Could someone find the name of the developer of this code? He's got some explaining to do, his relationship with computers and all...

  • All the comments variously raging against Microsoft/IE (totally unrelated to the topic), blaming the user for the browsers shortcommings, and finally the freetards telling him to fix it himself are all hilarious.

    I'd expect to read this on the linux haters blog, not on a random link from reddit. But good fun regardless.

    (Damn Yann, I guess someone else is seeing the hilarity)

  • People, please try K-Meleon browser. It is far ahead of Chrome, Opera and even Safari.

  • I didn't read past the third paragraph because this article is so atrociously written. Here's a little tip: adding more words doesn't make you sound smarter, it makes you sound pretentious.

    Consider making your point more concisely.

  • I'm using SRWare Iron. It's using the Chromium engine, just like K-Meleon is using the Gecko engine.

    I just tried out K-Meleon the first time and it's really worth a look.

    But for me the browsing experience is better in Iron, I like to have the tabs above the other stuff :-)

  • @Zonk: no-one forced you to read it. I tried to write an entertaining, yet informative, post, not a summation of dull facts. I guess, you prefer the latter, which is unfortunate in this case.

    @the people suggesting other browsers: we now have enough hints for other browser I think.

    @the people posting that their system isn't affected: luck you :) however others have this problem, so it's not that the problem doesn't exist, just that it doesn't affect everybody.

  • HELLo!

    Opera does indeed have an ad-blocker, they just call it "content blocker" as you can block anything with it. Right-click on the page, click "block content", and there you have it.
    To block content you can either just click on the images you want to block (holding shift if you only want to block a single image), or add the URI manually.

    Combine this with turning off plugins by default, which is configurable per page for those sites you want to have them enabled on, and you won't see an ad ever again. Well, unless you visit a new site with new ad-paths, of course. ;-)

    Happy codin'!

  • Switched to Chrome long ago, and FF3.5 is the new IE...

    It's a little sad, I'm all for open source, but this type of mistake doesn't really help.

  • I don't have this problem on my xp machine but I think this is because I seldom use IE or do anything else but browse on it and thus don't have that many temp files or cache files.

  • I was one that started using FF long ago when it was the begining of the FF era. But at work I had a slow PC with not enough memomory to share for web browsing. Then I made a discovery and that was the Opera. Since then I am trying to use FF but it is too hevy, too expencive for the memory, and it got slower. The new FF 3.5 was a promise for the speed but not for the momory consumption. Again is way beyond Opera or Chrome. And now you have this.

    So in time as popularity of the FF grows, the bugs are growing and by the time it wll become the same as IE. I like Opera because the stayed true for them selves. The Opera that I starded using and the Opera I am using now is the same. Has the same spirit, not too much features, only it is beatter and beatter.

  • I have an XP PC and have no problem with the version Firefox 3.5. It works pretty well very very very fast.
    I have read a guy on this forum who says "IE8 is a great offering.". Two solutions: he is working for Microsoft or he has never tried another browser !

  • completely disappointed. FF3.5 is slow, un-secure, tabs keep opening as new windows without even touching the mouse, right click on web sites does not exist. Luckily I have an FF3.0 installation saved on my hard drive, until they release the latest updates.

  • I too am experiencing this. I tried clearing all my cache/history from firefox but now I see why that won't work. It's sad but I also find myself gravitating towards chrome. It's a crappy user experience for developing (for me) but it gets better with each update (and when they get adblock equivalent).

    My whole point in liking chrome though is that when you close a tab, you can feel relieved that the memory actually does go away. Firefox has way better memory consumption with a lot of tabs open at once, but once you close the tabs, the memory doesn't go down. That's my biggest gripe. I hate it when my browser takes up almost 500MB of my 2GB of memory with only 1 tab. Forcing my XP to swap a lot and makes me into a task manager watcher (more so than I already am).

  • Minor point - The term is disk "thrashing". Trashing a disk involves corruption of files. Thrashing is where the disk read/write is going crazy and eats up all the resources.

    I have had a lot more issues with FF3.5 randomly crashing and momentary freezing. I've also lost a couple of plugins that I really miss. I'm not seeing a lot of reason to upgrade and a couple of reasons to roll back - mainly the 2 I mentioned.

  • I am using avast antivirus on WinXP and starting Firefox 3.5 is a task done while I go grab a cup of coffee. Every time I start Firefox you can see the antivirus, which is set to paranoia, go wild because of all the files accessed.
    On my desktop it isn't THAT bad, however, on my slow laptop with another program accessing the hard drive this can easily take 5 minutes.

    Also:
    Live was good. -> Life was good.

  • Startup time = 16 sec
    LinuxMint 7 (Ubuntu 9.04)
    CoreDuo 2.66

    Repeated many times - alway 16 sec.

  • I have been facing many issues mentioned here since the day i installed 3.5.. going back to 3.0.11

  • To AskAmyS:
    "randomly crashing and momentary freezing"

    FF 3.0 is also permanently crashing and momentary freezing, on Linux 64 bit. The problem is Flash / npviewer.bin. According my experience with FF 3.0:
    1. FF is randomly crashing: on Ubuntu 9.04 and Mint 7
    2. FF momentary freezing: on Ubuntu 9.04 and Mint 7
    3. System is completely frozen, even mouse doesn't move: on Ubuntu 9.04
    4. User session is terminated, but the system remains running, login dialog is displayed: on Ubuntu 9.04 and Mint 7
    5. System is halted (yes, immediately shut down on some pages with Flash): on Ubuntu 9.04

    But only the first two problems happen regularly, i.e. every 10-15 min on some web sites. The problems 3, 4 and 5 happen not so often, about 5 times monthly.

    Normally Flash means ads, so when I'm disbling Flash, FF 3 works perfect. But I'm still using FF 3, because other browsers are even more worse. And because 16 sec starting time of FF 3.5 is not acceptable.

  • It stopped working completly for me, maybe I'll try Opera now or Downgrade to Firefox 3, but I can't find the old version

  • I had the same problem too. I uninstalled FF then reinstalled it; nothing worked. I have no choice but to go to IE.

    Firefox is a joke.

  • For all we know, this could've been caused by a virus or spyware.

  • I do wonder who came up with the idea to read the IE temp folder. Very simple logic: if Firefox is the default browser (as hoped by Mozilla), then IE won't be used - meaning that the IE temp folder does not change. So, why use that as the seed?

    Stupid. I say. Truly stupid. I downgraded back to FF3.0 and now FF3.0 is slow too. Man, I've just formatted this PC. I give up with Firefox... I'm going back to Opera.. at least that one don't bug me almost every day to update some add-on and then bug me again after the add-ons are updated... Firefox is not the center of my world, you know...

  • Every since I downloaded Firefox 3.5 I have to refresh the screen after each page, they need to let us downgrade back to 3.0, because this totally sux! When will peeps learn, if it ain't broke don't try an fix it..its make for good old fashion common sense; besides wats so darn great about 3.5? Nuttin' n my book, lol.

    By the way, that Google Chrome sux too...I am in need of a completely new browsing adventure.

  • FF 3.0 can be downloaded here:
    http://www.mozilla.com/en-US/firefox/all-older.html

    FF 3.5.1 (with a patch for this problem) will be released soon, as there's a public exploit.

  • I have three xp intel double core PCs. And I confirm that the delay of Firefox 3.5 from strating is annoying: Almost 45 seconds. So it seems we cannot go back to previous Firefox versions...? It is nonesense. I need to migrate to Opera.

    None in Mozilla Org do test the new versions before launching to the the users?
    Pity

  • I got fedup and installed opera. Much better browsing experience. The 3.5 release feels half baked: not really ready, but we said it would be out by now so here it is....

  • This was indeed a legitimate issue, and has been fixed in Firefox 3.5.1, released today.

    Thanks for reporting it!

    — Alexander Limi · Mozilla User Experience

  • FF 3.5 starts in my computer in less than 3 sec. I've upgraded to 3.5.1 and I can hardly notice any improvement, though.

    Ubuntu 8.04 on a Pentium 4 3GHz:
    Add-ons: Adblock Plus, Adblock Plus: element hiding helper, British English dictionary, Deutsches Wörterbuch, Deutches Wörterbuch erweiter für Österreich, Diccionario Español/España, Download helper, Download Statusbar, Faviconize tab, FireFTP, Flagfox, Full Fullscreen, Google Preview, Lazarus, NoScript, Undo Closed Tabs Button,

  • The problem is that Firefox reads some temporary folders that if you use something different than windows XP/Vista/Win7 you probably won't have.

    I normally don't have any files in my temporary IE files so I don't have this problem. However, when my firefox slows down it means that my parents or siblings have used IE. >_>

  • BTW:
    on XP SP3 FF 3.5.1 refuses to open a new window (i. e via CTRL-N) if you have Download Helper (v. 4.6) installed.

  • I'm not a software person, so I have no idea why I have not been able to open Firefox over the past two weeks. My friends are having the same problem accessing FF.

    As a result, we have all switched over IE even though we have been loyal users of FF ever since its launch.

    I tried calling Mozilla headquarters, but they are not very customer friendly, so I have not been able to get through.

    That could be it for Mozilla. It's not hard to switch and we're all getting used to IE now.

  • I have downgraded back to Firefox 3.0 because I too found 3.5 to be a pain.
    It would crash overnight for three nights in a row.
    I leave my PC on all the time, but this never happened with 3.0.
    Also, it would not restore my last session the last time.
    So, I went back to 3.0 and so far so good.
    I'll wait a while, or until I have no choice to go to 3.5.
    And maybe I will even try IE8 or Chrome in the meantime.
    Too bad for Firefox. This is the kind of thing that generates loss of customer loyalty. They should have known better.

  • @Aleksey: Benchmarks don't mean much. What users "perceive" is much more important.

  • >I'm not going to advice friends ADVISE

    Hey grammernazi - should be GRAMMARnazi

Comments have been disabled for this content.