Slashdot Mirror
Is Firefox 1.0 Less Stable than Firefox PR1.0?
An anonymous reader wonders: "I had Firefox 1.0PR running smoothly on three different machines and it hardly ever crashed. After upgrading to 1.0, I seem to have at least one annoying crash a day. On one of the machines, using the 'self update' feature caused Firefox to crash in middle of the upgrade and left it in a completely unusable state. Eventually, I had to uninstall it and resort to using IE to download the full installer, again. Is it just me, or are other heavy Firefox users noticing this sort of behavior?"
Why don't you try posting on the Mozilla.org forums?
I don't know about anyone else, but I haven't had ANY issues with 1.0, perhaps the author of the article is using unsupported plugins / extensions that haven't been upgraded yet?
No bugs found here. But I was sad to see the 'Cookies are delicious delicacies' line disappear from Prefs.
[ UNSIGNED NOT NULL ]
I'm horrified that this is a front page post. What is wrong w/ you people?
Also, you are reporting the crash data back to the developers, right?
[o]_O
Welcome to Slashdot, now being used as an alternative for official software support sites and usenet newsgroups.
AT&ROFLMAO
Complaining is more fun than actually solving problems.
Mathematics is made of 50 percent formulas, 50 percent proofs, and 50 percent imagination.
I had the opposite occur. With 1.0PR I was having rather regular crashes when I opened the browser, and an annoying one that occured everytime I attempted to open any sort of streaming media. (I resorted to IE to watch the SpaceShip One launch).
However, with the full 1.0 I haven't had a single crash yet, and I've been using it a lot since the first day it was available.
This is not a sig.
Make sure you uninstall any old versions before installing the new version. Its in the faq.. well hidden, but i've had no problems when uninstalling then re-installing. Make sure to back up, but your savings should be saved as they are not held in the same area as the executeables and whatnot. I have had problems just upgrading, but i've been problemless since i've done the above.
"We are eternal.. all this pain is an illusion." -Maynard James Kenan
Go to CNN. View a few stories. **CRASH**
This is:
Linux 2.6, GNOME, 32-bit ppc, libswf installed,
multiple windows open, Debian-unstable, the tab
preferences extension installed so I can go back
to the old pre-tab Mozilla ways...
This really, really, sucks. I was one of those
people that would keep a browser running for
several weeks at a time. I'd let it sit on one
virtual desktop with two dozen windows open.
Mine seemed to be working but now it's showing an article about a single anonymous user's broswer problems on the front page of Slashdot. That can't be right. Hopefully Mozilla will have a patch available soon.
If you need to mod me down now, but I am starting to get PO'd!
Just what in the F*CK is going on with Slashdot???
Is there any justification what so freakin' ever this is a frontpage story? As far as anyone can tell this is about as informative and useful as 85% of the Usenet.
The quality of frontpage postings has gone down dramatically. After weeks of every story just being a heavily editorialiazed piece of crap, we now have, "Hey, does anyone elses FF 1.0 crash?".
Editors, Taco, Cowboy Neal?!? Is anyone awake here? Have we totally lost our standards?
Cripes.
I left the lights on in my car the other day and then my car wouldn't start. Could these two events be related? Did anyone else ever have this happen?
word.
I've had random crashes browsing MapQuest. It seems around one out of every ten times I refresh a map, Firefox goes boom. I actually suspect it may be some Javascript/Java/DHTML/etc. in an advertisement that's in their rotation.
-Z
While I'm not going to disagree with you, I fail to see how this absolves the developer's responsibility to build the application such that it doesn't try to absorb all the system's resources in the first place.
I'm currently using FireFox PR1.0, it's been open for about 5 days straight, running on Win2KPro. It's using 104MB of RAM. Why I don't know... I only have 4 tabs open at the moment and no flash or java running, and no third party plugins... but it's using 104MB or memory right now. It probably would have locked up if I didn't have a gig of RAM in this machine...
Went up to 108MB when I hit preview, and it's not going down...
=Smidge=
Interpreting how much memory an application is using is somewhat difficult. Modern virtual memory operating systems page memory into both physical RAM and disk storage (swap, pagefile, whatever). Next, there are operating system features that try to preserve actual memory by not allocating real chunks of memory on an application's request, until that memory is really made use of. Finally, using some system GUI widgets etc can "increase" the amount of memory use as shown in Task Manager etc where really the memory use is within the OS, not the app.
:)
What this comes down to is: the figure you generally see for memory use of an app is not physical RAM use. It might not even reflect the actual amount of physical+disk memory in use! Finally, memory usage might be overstated due to transient external allocations (e.g. win32 API dialog boxes) that deceivingly appear as memory used by an application.
What you have to look for is how that memory usage figure changes over time. In most cases, it grows until it hits a ceiling - even at that point, it is way overstated (a conservative measure, so to speak). What is bad is if it regularly grows by 50 MB per day, without limit. Then there is a leak
My coworker and I just upgraded to 1.0 and have bothed noticed that firefox will eventually eat up most/all system memory, forcing us to kill firefox and restart it. (we ran the beta/pre releases with no issues) The only common factor is that we both had to reinstall the web developer (version .8?) after upgrading to FF1.0....
Could this problem be related to specific extensions and not specifically to firefox itself?
it's been open for about 5 days straight, running on Win2KPro. It's using 104MB of RAM.
Even worse, there's this System Idle process that's taking up 99% of my CPU time!
Sheesh. It's called memory caching. That's why TOP differentiates between RSIZE, VSIZE and RSHRD.
RSIZE is the amount of ram being actively used by a process. I doubt RSIZE is 104megs.
1.0PR had a javascript pop-up crash bug that drove me crazy. 1.0 fixed that.
Some things to consider:
1. How did you install 1.0? Did you do an overwrite? If so, do a clean install.
2. What extensions are you using? Have you disabled the extension version check?
>On one of the machines, using the 'self update' feature caused Firefox to crash in middle of the upgrade
When was this? Do you have DNS/network/firewall issues which could be causing this?
Lastly, to get some real answers from the experts people should asking here.
1) Why is your faith in open source so great that you are unwilling to humor the idea that there is a bug in the application? I am a professional software developer, and most of us LAUGH at the idea of "bug free" applications. It is considered, in some circles, so laughable, that one is considered a bit of a neonate to tout that their software is bug-free [that, or not connected to reality (at least, with respect to the technology)]. The first time I ran Linux (late 90's) I had plenty of core dumps.
/. not Coast to Coast AM.
2) Are you proposing that Microsoft has a "black ops" department, whose sole purpose is to cause Windows to behave incorrectly when 3rd party software is run? Additionally, this department is exceptionally good at keeping a secret. So good, in fact, that the only way to detect their work is by running open source (patently bug-free) software on their OS, to uncover these flaws? Given that open source software is bug-free, wouldn't such a department fear discovery when performing such an act?
I'm not saying that there aren't reasons to dislike Microsoft, but goodness, this is
I'll tell you the single source of all of your Microsoft woes... the market. If the market will pump billions of dollars into a company, they have little right to complain about that company's software. There is competition. There was a lot more of it before all of you gave them all of your money. If you dislike Microsoft's product line, then download a Linux or BSD ISO, and install it. If you vote with your pocketbook, the company will listen. Hit companies that break the law with the law, and if you dislike the lack of competition, then purchase a competing product, or compete with them.
do you leave firefox in a page with flash animations? nowadays many of the animated figures in a web page are flash and not animated gifs anymore. a badly made flash can take up tons of CPU even when you leave it there. also, are you viewing web sites with automatic page refreshing? it will eat up the CPU everytime the page refreshes.
if that bothers you, you can always use the task manager to set the process's priority to either "below normal" or "low".
however, games are memory intensive. so as a browser, which uses memory caching to be fast. when real memory is used up, "thrashing" occurs (to swap some memory pages to the disk). even adjusting task priority won't help here, since thrashing is inherently slow. whenever a web page that you leave in the background refreshes itself, the OS has to swap out a few pages of game memory and swap in memory pages for the browser. as the game continues, it needs the memory back, and the OS has to juggle around memory pages again.
if you see a periodic frame loss, then self-refreshing web pages are definitely the culprit.
I once had a signature.
Open source software is not bugfree by any means. But the development time is by orders of magnitude faster than that of closed source development. Combine that with lots of eyes and there are no "easy" bugs left like there often are in closed source offerings.
Most closed source supporters don't even dispute this. Instead they claim that it only works in a very popular project and most projects arent. There is a counter argument to that but I won't go into it. The point is that firefox is an extremely popular project so that argument doesn't refute that the open source development model should yield the ideal result.
Microsoft probably has the worst reputation for stability in the software industry and it is one that is not lacking in actual merit by any account. So this is not exactly random piece of open source software versus random equivelent closed source software.
This is one of the best open source applications with repute for one of the cleanest designs, against the absolute worst repute closed source firm and further a product of theirs with a known terrible design.
"Are you proposing that Microsoft has a "black ops" department, whose sole purpose is to cause Windows to behave incorrectly when 3rd party software is run?"
You try to make it sound like he is mel gibson in a Conspiracy Theory. This isn't exactly that, this is a company with a PROVEN track record of doing just that, RECENTLY and doing so IN THE BROWSER MARKET specifically. Read up on the MSN website and the way it renders in Opera.
Further you speak as if your a developer on windows, that means if you've done anything non-trivial you know that windows does NOT behave as documented in NUMEROUS cases. Finding a system call which is only used by one competitor with a significant user base and making an intentional design change to break them is easy enough. If too many people report the problem you can claim it's a bug introduced by the update instead of an intentional change. This would be blatantly obvious if the relevant source were revealed, but it's not.
It's hardly a conspiracy theory at all to believe Microsoft would engage in any illegal anti-competative practices they can which they believe will ultimately preserve more market share than they will cost.
Particually after US CERT advisories to change browsers firefox has become a serious threat to their browser monopoly. I'd venture Microsoft would be willing to risk a substantial number of customers to discredit the stability of the first fully stable release of the first significant threat to their monopoly in almost a decade.
"If you vote with your pocketbook, the company will listen."
Well, either listen, or engage in anti-competitive practices to ensure you no longer have that option. After all, allowing you the option to choose a competitors product simply because it is superior or you don't them is a bad business strategy! First step is as simple as a minor api change that doesn't affect many applications but that firefox uses. This way you can make some customers feel firefox is unstable. That way you can buy time until you can get DRM'd which use encryption that is only compatible with windows. After all, you've already got the DMCA in place to ensure competition can't beat that simply because they figure out how the encryption works.
Then you don't have to worry about competiton on inexpensive x86 systems anymore.
I'm going to guess that over the five days, you have opened and closed a whole bunch of tabs (probably dozens). It's a known issue in Firefox that when you close tabs, it doesnt release the memory.
See the bugzilla: bugzilla.mozilla.org/show_bug.cgi?id=131456
I've had FF running for a week straight and using upwards of 200MB and only one tab open :). The only remedy is to restart FF. This has been an issue for over two years now. Dont expect it change anytime soon though, if it was a simple fix, I supppose it would have been done by now.
I've been experiencing the exact same phenomenon. So have a few of my friends. I'm sure it's not happening to everybody, but yeah, for me, the PR seemed more stable. On my system the official release goes to 99% CPU utilization and has to be shut down a few times a day, typically.
It doesn't throw errors to report. I'm not savvy enough to know how to get debugging information out of it, and I don't have the time to spend on mozilla forums trying to get someone's attention and then working it out.
So I won't put in the time. I don't expect the firefox people to fix it for me, given that, of course. They've already given me plenty, and it's still a great browser.
But I have been having this problem, and if other people have, too, then I'm glad to see it being discussed. Beyond hoping the problem becomes well-characterized, I think it's worth having a discussion about this because it could have implications for how OSS is perceived by the mass culture. The Firefox campaign is the biggest, most successful open source push in recent memory. Let's not act like it's heresy to talk about it here instead of in a newsgroup somewhere.
Try uninstalling Firefox 1.0PR, don't worry the profile will be left intact, then install Firefox 1.0, this works just fine, and Firefox 1.0 is completely stable.
"I had to uninstall it and resort to using IE to download the full installer, again."
What, you don't know how to ftp from the prompt?
sig not found
Dude, we're not talking impossible here. I'm typing this from a Linux notebook, and I own only the Microsoft products that came with my equipment.
/. really believed that strongly in the movement, then they would do what I do and run it on their home equipment, and we wouldn't be worried about Windows bugs affecting open source products.
The point of my argument was that it's absolutely annoying to no end to hear people claim that just because a piece of software is open source, it can't be that software's fault. Certainly at some point in ones life, they have to accept the things that they like as "good" without needing to consider it "flawless."
I am an open source advocate, but I feel that making laughable claims in support of open source software is no way to promote it. How about, before we go pointing fingers, we take a quick look at the problem, and then prove that it's Microsoft's fault.
If open source software were naturally bug-free, nobody would be running software to track those bugs. If everyone in open source were a developer, there would be many more developers here.
If everyone on