What is Digg?Sponsored by Rockstar Games
Grand Theft Auto: Episodes from Liberty City view!
rockstargames.com - Out Now on Disc for Xbox 360. Includes The Lost and Damned plus the all-new The Ballad of Gay Tony.
343 Comments
- breaknenter, on 03/12/2008, -3/+419The guy who wrote this is one of the lead engineers for Firefox. Good article if you're interested in what they did to fix the memory problems in Firefox.
- honkerdown, on 03/12/2008, -4/+205I can't wait until a few more of my extensions have Firefox 3 suport.
- Shao00, on 03/12/2008, -5/+205using 3 beta 4 now, its been awesome so far.
- jaxter2010, on 06/17/2009, -11/+176I try to not succumb to fanboyism, but firefox is just awesome.....
- inactive, on 03/12/2008, -23/+122It's so interesting that they're fixing leaks that they said didn't exist and reducing excessive memory usage that we were told was just a figment of our collective imaginations.
- Dyar, on 03/12/2008, -17/+101I am getting a different graph for IE7.
How I measured the performance on windows:
Start->Run...->perfmon
Delete all the counters.
Add Counter...
Select your "iexplorer" process under Performance Object
Select the "Private Bytes" counter as suggested in the article and click Add
You may need to change your graph to have the counter value appear in the middle. I changed the max value of my graph to be 2000
In your IE7 open new tabs and browse to different sites
Observe as your memory usage go up. Wait until it reaches a steady level.
Then close your tabs until you are left with a blank one.
Observe the memory usage again.
For my graph the usage goes back down as the tabs close. I can't say if it goes back down to exactly the same level as it started but it definately goes down. Not the same as the graph shown in the article.
Instead of believe everything they read (including what I just said), I suggest people try it themselves. - inactive, on 03/12/2008, -8/+56This is flawed. I was just arguing about about this with my roomate / classmate. Perfmon probably gives a better view than taskmanager but IE7 somehow masks the memory it is actually using. Try this test now as well:
Open up IE7 and open 15 tabs with different websites. Simple websites will show around 97,000 while a complex website will show a completely different number for memory usage when you move the tab over. Firefox on the other hand, always reports the same amount. I dont know where the memory is going, but IE7 masks it somehow, so using any windows based tool to measure IE7's memory usage will always produce a flawed result. - Tyr7BE, on 03/12/2008, -2/+40http://mozillalinks.org/wp/2008/02/firefox-3-ultim ...
"Along with the long list of new features and improvements we can find in Firefox 3 betas so far, there is a series of performance improvements coming to Firefox 3 Beta 4."
Looks like it whupped Opera's ass in the javascript test, which is what accounts for most of the stalls and slowness you see while browsing the web. - comrade693, on 03/12/2008, -5/+42It's quite a common misconception that they've denied that they leaked...
All large software applications leak. People were pointing to increased memory usage as leak, which isn't a strong indication of leaking. The blog post that everyone always refers to said that the increased memory usage wasn't a leak, but that it was because of a feature (which was true too!). - nights0223, on 03/12/2008, -4/+41Just use Nightly Tester Tools. It will make all of your extensions compatible.
- dcd722, on 03/12/2008, -5/+38The damn dog shows up everywhere. And I still can't shoot him =[
- pcabellor, on 03/12/2008, -1/+33comment by Mozilla's Mike Schroepfer:
Since folks asked I ran the latest Opera 9.5b in the exact same environment. It peaks around 240MB and doesn’t free up any memory at the end (so ends at 240MB). Performance during the run is similar to Firefox 2.0.0.12 but higher than Firefox 2.0.0.12 at the end. It is significantly higher than Firefox 3 - which peaks around 220MB and ends at 85MB. I sent stuart the raw data if he wants to update the graph.
http://blog.pavlov.net/2008/03/11/firefox-3-memory ... - Ramble, on 03/12/2008, -15/+46And why did they not include Opera 9.5 beta?
- lucidapathy, on 03/12/2008, -2/+32TINYURL rickroll... some people have way too much ***** time on their hands.
- LiquidIse, on 03/12/2008, -2/+31I am liking this trend in the standard-compliant browsers of hard-hitting performance increases. I persoanlly am using a recent nightly of webkit, and it is worlds ahead of FireFox2 or Safari3. Anxious to see the official release of both next-gens.
- epohs, on 03/12/2008, -0/+27Definitely feels faster (and also more stable) than 2.0.0.12
- inactive, on 03/12/2008, -0/+25gotta disagree with Avid...
Firefox 3B4 on both Windows XP and Ubuntu Gutsy have very noticeable performance improvements. Plus, the address bar changes are addicting once you get used to it, and the built-in download manager works a lot better. Haven't noticed any bugs at all really... even though it's "beta" - br0ck, on 03/12/2008, -0/+23Actually, I can now actually open a number of larger Digg comments pages simultaneously on an older machine. Before this beta, the entire browser would hang for up to a minute, but now all of the comments pages load smoothly in the background while leaving the rest of the browser responsive and usable.
- sppavlov, on 03/12/2008, -3/+26Graph including Opera and one of the Safari runs that crash:
http://flickr.com/photos/stuartp/2328802961/ - humbled, on 03/12/2008, -3/+25Not necessarily so. The problem is the conflation of committed and actually consumed bytes. In the chart, he is showing the bytes that are actually used by the app. If IE tries to commit 100 megabytes, then uses 60, and then releases 20 of the committed but not needed 40 remaining megabytes, you'll see a 20 megabyte drop on your XP box even though IE has not actually freed any of its memory. Your results are consistent with the article, they do not refute it.
- InfiniteNothing, on 03/12/2008, -2/+24Vista? they mentioned that private bytes were not quite correct in XP.
- adrianmonk, on 03/12/2008, -5/+26I'm assuming you mean it will disable the checks for whether they're known to be compatible.
Obviously, without some extremely amazing artificial intelligence, it probably can't really make them compatible. Also, those checks are there for a reason. It's legit to bypass them and try it out, but once you do that, there are no guarantees. - lostboy, on 03/12/2008, -0/+20I actually got frustrated because I tried the beta and found the new bookmarking engine Places so great that I didn't want to go back to ff2.
- goblindegook, on 03/12/2008, -0/+20You don't need that anymore, just go into about:config and set the extensions.checkCompatibility flag to false (you may need to create it first). As the poster above me said, there's no guarantee the extensions will work and may even screw up your browser configuration.
- gcauthon, on 03/12/2008, -2/+22They're not so dumb that they would run a performance test with a debug-checked build.
- epohs, on 03/12/2008, -0/+20Agreed. This is a good overview.
I really hope they continue to focus on increased performance and better memory usage. Those kinds of things may not get as much attention as a snazzy new feature, but they are so very very appreciated... even if each individual tweak doesn't get the applause that it probably deserves. - trisweb, on 03/12/2008, -0/+19Oh, but rendering speed has increased as well, by a noticeable and measurable amount. Best of both worlds :)
- anillop, on 03/12/2008, -19/+38Because no one cares about Opera, and the last thing firefox wants to do is give free press to a competitor. They both fill a small niche of people who are willing or tech savy enough to turn away from IE.
- saisumimen, on 03/12/2008, -2/+20They're talking about browsers normal people use ;p
- luke16, on 03/12/2008, -3/+21Been using firefox 2 since release, shall i just switch to 3 beta?
- trogdoor, on 03/12/2008, -1/+18/me makes sure his "hello world" firefox plugin is compatible with FF3...
- mrdoogee, on 03/12/2008, -0/+16I'll have you know people still use Netscape 7 here where I work.
Yes I'm IT
Yes I hate them. - taketheleap, on 03/12/2008, -3/+19speak for yourself... FIREFOX FTW!!!!
wheeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeee! - jtinz, on 03/12/2008, -1/+17Mozilla has offered free T-shirts to the authors of extensions if they make them compatible with Firefox 3.x until March 18th. I think more than a few authors will try to meet the deadline. You cannot buy those T-shirts.
- GramarNatzi, on 03/12/2008, -15/+29Because opera wipes the floor with firefox when it comes to memory usage. Don't get me wrong, I like firefox and use it on my desktop all the time. But on my EeePC firefox burns up too many of the limited resources available and becomes sluggish unresponsive. Opera on the other hand runs great.
- inactive, on 03/12/2008, -4/+18Bitches don't know about my extensions.
- trisweb, on 03/12/2008, -0/+13Heck yes it's faster! Amazing how big a difference it makes...
- dunbone, on 03/12/2008, -4/+17Firefox runs at a higher CPU percentage than IE7 and thus chews up batteries quicker in laptops. No kidding, I've seen over 250 cases of this first hand
- digjam, on 03/12/2008, -1/+14I will pretend that I understood all that and say ..BRAVO!
- Nanobe, on 03/12/2008, -0/+13I've been using the Firefox 3 nightly builds for a month or so. In my experience, the nightly builds are actually more stable than the Firefox 2 final release (at least on Linux). The only crashes I have experienced have been from Flash, but that's just a result of Adobe Flash's general suckiness on Linux.
I have grown so appreciative of the "awesomebar" that I don't even miss my not-yet-compatible extensions that much. - goblindegook, on 03/12/2008, -0/+12Significantly, yes. I've been using it on my old Pentium 4 laptop and I can certainly tell the difference.
- comrade693, on 03/12/2008, -3/+14Oh, I see what you did here. You made a point that is only slightly related to the issue (memory leaks) but not actually addressing it to argue against my claim to make yourself look right and me wrong. Fascinating!
- pagefaultca, on 03/12/2008, -0/+10Firefox is not ACPI aware and does not throttle itself. This is the main problem.
- jakethecake, on 03/12/2008, -0/+10New Graph -> http://flickr.com/photos/stuartp/2328802961/
Posted via Pavlov's flickr widget.
It includes Opera 9.5 and Safari 3.0.4. - f3l1x, on 03/12/2008, -9/+19ITS BETA... debugging symbols most likely not stripped, logging for devs, bugtracking, etc.. Also, Firefox doesn't piggyback as much resources off the OS as IE does.
- serialvibe, on 03/12/2008, -2/+11How is that a bias?? Breaknenter stated that it's an article by a lead engineer, and that it's a good article for those interested.
- Jaliyl, on 03/12/2008, -1/+10You win Funniest comment of March 12 2008.
- Pegritz, on 03/12/2008, -5/+14I absolutely love Firefox 3. Just a few months ago, I had abandoned FF 2 for Opera (which I've never, ever liked, but the only other alternatives were Flock, which is just FF 2 with a different skin, and IE7, which crashes on me constantly). I'd heard FF 3 was supposed to deal better with memory issues, but I wasn't about to believe it until I saw it, so the second Beta 1 became available I downloaded it to see if ANYthing could be an improvement over that memory-hog piece of ***** FF2.
I'm absolutely amazed. It works like a charm. I'll have to give the FF programmers credit: they said they'd reduce memory usage and, goddamn, they sure have! - trisweb, on 03/12/2008, -0/+9Nope.
But isn't it strange that despite the ever increasing number of easy fast food choices and large-person-accessible homes and furniture, we're still hung up on fitness and health and being thin?
Any programmer that ignores memory efficiency is greedy and lazy. It is always true that [a small, fast, efficient program] > [a large, slow, greedy program] given equal functionality, and therefore wasting memory is not ever good.
Besides, program obesity is just a sign of larger problems, and will lead to larger problems. Better to be lean and lightweight. - Dyar, on 03/12/2008, -3/+12Interesting. I'll try to verify this with a ThirdParty profiler.
However I don't see how someone can monitor IE7's memory usage unless it is a windows base software since IE7 only runs on windows. Even if you try to remotely do it, you will be executing WMI queries which basically is the same as what is shown in perfmon.
Either way thanks for heads up. I definately won't say that IE7 uses less memory than FireFox but I also can't imagine even Microsoft would release a product which does not free the memory it no longer uses. -
Show 51 - 100 of 346 discussions
© Digg Inc. 2009
— Content created and posted by Digg users is