I doubt it was extensions, I only run GMail Notifier and AdBlock.
AdBlock did have severe memory leaks, and may still leak quite a bit of memory. Have you upgraded to the very latest version? All it takes is one bad extension to leak memory like a sieve.
How does what the size of the version increment deteremine how bloated or bug ridden software is? Anyway, Firefox 2 was available as a "preview" as soon as Gecko 1.8 branched off the trunk last year. The day after any stable branch is created, a new trunk build is produced that is the first preview of the next stable version. Is there some kind of problem with doing that?
Nice. So call it something, release it into the public (as is required by the license) and insist it doesn't exist. For a product to bear a new name means it is newly branded product, thus a nerly branded product has been released.
So you would rather they not rebrand the browser until the very last patch, at which time they are completely committed to release it to the public without testing? Frankly, I don't see the point. Why not rebrand the product and allow it to be fully tested before release?
All browsers have lots of memory issues. They also all have security problems, they crash under lots of different situations, have many kinds of CPU use problems, and thousands of other kinds of bugs. What else is new?
I don't think embedding a database will noticably impact memory usage. The most noticable change will be that your bookmarks, cache, and other parts of your profile will not be corrupted or lost nearly as easily. The dataloss problems SQLite will fix are much more severe than the memory problems some people expereience with Firefox. Firefox is using too much memory? Just restart. Firefox lost your bookmarks? Tough sh*t!
And yet in the previous story, users are seemingly rushing off to upgrade to Firefox 2.0 alpha. Sometimes you really do need to state the obvious.
At least wait for the.0 versions if you don't want problems, folks. You might want to wait for the.1 or.2 versions. This applies to appliances, cars, software, and even books (I try to wait for the first corrected printing for O'Reilly books).
1. Mozilla puts the release out so users can report bugs. Unless you already have a Bugzilla account, I suggest you wait until a release candidate is out.
2. If Firefox 1.5.0.1 is crashing so often or leaking memory so badly for you that you need to restart Firefox every day or so, you might want to try 2.0 Alpha to see if it fixes your problems. Of course, if it doesn't you should report the problem (see #1).
End users generally should not be downloading alphas because of the downsides you mention.
Unless you're fairly deeply involved in QA (i.e. you have a Bugzilla account and plan to report bugs you find), just skip alpha versions and stick with a stable release. Wait until the beta or a release candidate is out much later this year.
Nearly all browser stats programs detect Opera properly, even when cloaked as IE. Even Chuck Upsdell, who compensates for the fact that Opera is likely undercounted, puts Opera usage at around 1% and Mozilla browser usage around 11%.
So for anyone who did misread the headline and thought you finally had complete triumph over all the creationist wackos, I hate to burst your bubble.
The debate about evolution vs. creationism has nothing to do with which is right and which is wrong. It's about creationism not being based upon science, but based upon faith, and as such it has no place being discussed in a science classroom.
I don't think that increased diversity has made it more time consuming to develop web applications. Back when Netscape 4 and IE 4 were the big two browsers, it was awful to write sites that worked well in both browsers. Now that we have IE, Firefox, Safari, and Opera, even though there are more browsers to potentially test with and have issues with, it's much easier to write sites that work in all browsers because they support the standards so much better than the older browsers. Usually all I need to do these days is make a site work in Firefox and validate the HTML and CSS, then tweak it for IE a bit, and it tends to work well in all browsers.
Microsoft doesn't make a Flash plugin that you need for Flash sites, or a PDF reader that you need to view PDFs. Why do they need to make the browser that you need to view web pages?
Also doesnt suffer from the memory leak problems that "dont" exist in firefox.
Where do you trolls come up with this stuff? Of course every version of Firefox has memory leak bugs in it. So does every browser. And every browser crashes, and has security problems, and hangs and has other CPU problems. Get over it.
If I am just a user, why should I really care about standards ? No seriously why should I ?
So stuff works without you having to think about it. If web developers write a site that works in IE only, and you want to go to that site with any other browser, it won't work. If you happen to be on a Mac or Linux computer, you don't even have the luxury of using IE. It just won't work. If everyone follows standards, things just work.
It's exactly the same principle as other standards, such as the standard rail gauge that allows standard trains to ride on all standard tracks. Do you want websites to just work? If so, you should care about standards.
One day, I accidentaly left Fx running since morning, with just Slashdot open in one tab. When I got home that night, Fx was up to 870 megabytes of memory usage
Clearly, that's an extreme memory leak, one of the worst examples I've heard of. It's obviously a major bug, and you should report how to reproduce it so it will get fixed. No one has ever said that that degree of memory usage is caused by the Back-Forward cache feature, as you're trying to claim.
No one said memory leaks were caused by a feature. Ben Goodger explained that the obvious increased memory use of Firefox 1.5 was casued by the Back-Forward cache feature. He also stated explictly that all versions of Firefox leak memory -- and of course memory leaks are bugs, not features.
The Back-Forward cache causes immediate increased use of memory, just after loading a few pages. The increased memory due to memory leaks doesn't become apparent until after visiting hundreds of pages and several DOM Windows have leaked. That's why he said that the increased memory use people were complaining about is a feature, not a bug.
I hope the difference between the Back-Forward cache (a feature) and memory leaks (bugs) is now clear. Just because both cause Firefox 1.5 to use more memory does not mean both are bad. The feature is good, and the bugs are bad.
No one is lying, except possibly you. Enjoy Opera, the browser of whiners.
The "monkeys" at Mozilla are well aware there are memory leaks in Firefox. That's why they developed the leak-gauge tool to help find memory leaks. I'm using the leak tool, and I can see the latest nightly build of Firefox 1.5.x still leaks 1% or more of the DOM Windows it creates, and a leak of that severity could easily cause memory usage to increase by hundreds of megabytes over the course of many days.
I agree that the download figures are meaningless.
My take on download figures is that they provide a very rough estimate of the number of users who have tried a product. In this case, Firefox has been tried by roughly 150 million users. Perhaps the number is as high as 200 million, perhaps as low as 100 million, but it seems unlikely that it's been tried by as many as 400 million or as few as 50 million.
When you add that there are about 1 billion Internet users and Firefox has about 10% usage share, there seem to be roughly 100 million current Firefox users. Then we can say that it seems likely that most users who have tried Firefox are still using it.
Instead of these numbers being meaningless, it seems like Firefox is one of the most downloaded programs ever, and most users who try it are satisfied with it. Perhaps only WinZip has had more downloads. It might be that Firefox is the most successful open source product ever. I don't see that it's misleading or immature to boast about these facts.
I'm referring to the Forrester Research report described in the linked article. Some people have read that article and come to the same conclusion you have — that web stats are so completely unreliable as to be useless. If you read the article carefully, you'll find it says no such thing. It merely states that stats from web analytics firms are not as accurate as sometimes claimed. You can see from comparing stats from different sources that reported Firefox usage may vary by up to 30-40%, but I can't see any evidence that the stats figures are a factor of two off, much less an order of magnitude off as you seem to suggest.
If you take the Firefox usage share data from various stats sources as a sample of the actual usage share of Firefox, you'll find that Firefox usage is 10.1% +/- 1.5%. This means there's a 95% chance Firefox usage is between 7% and 13%. That makes it very likely that there are between 70 million and 130 million Firefox users. An easy way to say that is that there are about 100 million Firefox users.
I never said web stats are 100% accurate. Of course there are some inaccuracies in web stats. One recent study found up to a 40% discrepancy between two web stats sources. However, it's one thing to say they're somewhat inaccurate, and quite another to say they're "full of it".
It still seems to me there are about 100 million Firefox users, as opposed to 200 million or 50 million. If you have a better estimate, let us know what it is and the data and techniques you used to arrive at it.
Slashdot Mirror
Close, but not quite right. Firefix 1.5 uses Gecko 1.8.0 and Firefox 1.5.0.1 uses Gecko 1.8.0.1. Firefox 2.0 will use Gecko 1.8.1.
New features planned for Firefox 2.
There's a new reflow factoring branch off the trunk where the work for Acid2 is taking place.
How does what the size of the version increment deteremine how bloated or bug ridden software is? Anyway, Firefox 2 was available as a "preview" as soon as Gecko 1.8 branched off the trunk last year. The day after any stable branch is created, a new trunk build is produced that is the first preview of the next stable version. Is there some kind of problem with doing that?
The problem described looks like Bug 259672.
I don't think embedding a database will noticably impact memory usage. The most noticable change will be that your bookmarks, cache, and other parts of your profile will not be corrupted or lost nearly as easily. The dataloss problems SQLite will fix are much more severe than the memory problems some people expereience with Firefox. Firefox is using too much memory? Just restart. Firefox lost your bookmarks? Tough sh*t!
At least wait for the .0 versions if you don't want problems, folks. You might want to wait for the .1 or .2 versions. This applies to appliances, cars, software, and even books (I try to wait for the first corrected printing for O'Reilly books).
1. Mozilla puts the release out so users can report bugs. Unless you already have a Bugzilla account, I suggest you wait until a release candidate is out.
2. If Firefox 1.5.0.1 is crashing so often or leaking memory so badly for you that you need to restart Firefox every day or so, you might want to try 2.0 Alpha to see if it fixes your problems. Of course, if it doesn't you should report the problem (see #1).
End users generally should not be downloading alphas because of the downsides you mention.
Unless you're fairly deeply involved in QA (i.e. you have a Bugzilla account and plan to report bugs you find), just skip alpha versions and stick with a stable release. Wait until the beta or a release candidate is out much later this year.
Nearly all browser stats programs detect Opera properly, even when cloaked as IE. Even Chuck Upsdell, who compensates for the fact that Opera is likely undercounted, puts Opera usage at around 1% and Mozilla browser usage around 11%.
Here you go!
I don't think that increased diversity has made it more time consuming to develop web applications. Back when Netscape 4 and IE 4 were the big two browsers, it was awful to write sites that worked well in both browsers. Now that we have IE, Firefox, Safari, and Opera, even though there are more browsers to potentially test with and have issues with, it's much easier to write sites that work in all browsers because they support the standards so much better than the older browsers. Usually all I need to do these days is make a site work in Firefox and validate the HTML and CSS, then tweak it for IE a bit, and it tends to work well in all browsers.
Microsoft doesn't make a Flash plugin that you need for Flash sites, or a PDF reader that you need to view PDFs. Why do they need to make the browser that you need to view web pages?
It's exactly the same principle as other standards, such as the standard rail gauge that allows standard trains to ride on all standard tracks. Do you want websites to just work? If so, you should care about standards.
The Back-Forward cache causes immediate increased use of memory, just after loading a few pages. The increased memory due to memory leaks doesn't become apparent until after visiting hundreds of pages and several DOM Windows have leaked. That's why he said that the increased memory use people were complaining about is a feature, not a bug.
I hope the difference between the Back-Forward cache (a feature) and memory leaks (bugs) is now clear. Just because both cause Firefox 1.5 to use more memory does not mean both are bad. The feature is good, and the bugs are bad.
No one is lying, except possibly you. Enjoy Opera, the browser of whiners.
No one is denying that there are memory leaks. However, they're not common (occuring on only about 1% of visited pages) and often very hard to reproduce reliably. You can help by using the memory leak tool and reporting good memory leak bugs.
When you add that there are about 1 billion Internet users and Firefox has about 10% usage share, there seem to be roughly 100 million current Firefox users. Then we can say that it seems likely that most users who have tried Firefox are still using it.
Instead of these numbers being meaningless, it seems like Firefox is one of the most downloaded programs ever, and most users who try it are satisfied with it. Perhaps only WinZip has had more downloads. It might be that Firefox is the most successful open source product ever. I don't see that it's misleading or immature to boast about these facts.
If you take the Firefox usage share data from various stats sources as a sample of the actual usage share of Firefox, you'll find that Firefox usage is 10.1% +/- 1.5%. This means there's a 95% chance Firefox usage is between 7% and 13%. That makes it very likely that there are between 70 million and 130 million Firefox users. An easy way to say that is that there are about 100 million Firefox users.
It still seems to me there are about 100 million Firefox users, as opposed to 200 million or 50 million. If you have a better estimate, let us know what it is and the data and techniques you used to arrive at it.