Good news: Mozilla fixed their damn browser • Hacker News Board

Re: Good news: Mozilla fixed their damn browser

#31

Post by bzbarsky » Tue, Dec 20, 2011, 11:55 PM UTC

CrLf wrote: post #7

jmillikin wrote: post #4

Not to discount the work that Mozilla has been doing, but it seems like the "improvement" they added is actually just a workaround for distros that configure filesystems incorrectly. That is, his machine is configured to use an obsolete filesystem (ext3) in a mode with significant performance problems (data=ordered). There are numerous ways that a machine can be mis-configured such that it suffers from poor performan…

I find these statements amusing... As it happens, neither is ext3 an obsolete filesystem (I have many machines using it just fine, thank you, with no intention to migrate to less tested filesystems just because they're "new"), nor "data=ordered" has significant performance problems. More amusing than this just the fact that the filesystem is a bottleneck for a web browser... an application that should be network and…

Browsers are CPU-bound, network-bound, GPU-bound, memory-bound (esp. memory bandwidth, though some people count that as "CPU"), and filesystem-bound. Depending on exactly what you're doing with the browser.

Re: Good news: Mozilla fixed their damn browser

#32

Post by otherpope » Wed, Dec 21, 2011, 12:05 AM UTC

kakuri wrote: post #25

NARKOZ wrote: post #16

Earlier quoted context omitted.

Use Chromium if you complain about privacy

SRWare distributes a Chromium build called Iron that strips out privacy problems. http://www.srware.net/en/software_srware_iron.php

I used to recommend Iron but I don't anymore. I became increasingly uncomfortable with the lack of documentation and lack of transparency in the project. They delivered their source code as two code dumps on rapidshare last time I checked. Something just doesn't feel right at all.

Re: Good news: Mozilla fixed their damn browser

#33

Post by viraptor » Wed, Dec 21, 2011, 1:13 AM UTC

joejohnson wrote: post #19

What are people's overall impressions of Firefox 9?

I don't remember anymore - running 11.0a now. I'm using nightly builds since ~2009 and haven't experienced any issues so far (really - not even a random crash). The difference is so huge in overall performance, I don't want to go to the "stable" release anymore.

The only problem are some less common plugins, but adblock and firebug get updated within days... and I just switch to the latest official release for quake live.

Re: Good news: Mozilla fixed their damn browser

#34

Post by sciurus » Wed, Dec 21, 2011, 2:09 AM UTC

CPlatypus wrote: post #15

CrLf wrote: post #7

Earlier quoted context omitted.

I find these statements amusing... As it happens, neither is ext3 an obsolete filesystem (I have many machines using it just fine, thank you, with no intention to migrate to less tested filesystems just because they're "new"), nor "data=ordered" has significant performance problems. More amusing than this just the fact that the filesystem is a bottleneck for a web browser... an application that should be network and…

The problem with data=ordered is that it imposes a global order, allowing one I/O stream to stall others. Imagine if waiting for data to be acknowleged on one TCP socket meant waiting for all data sent earlier on all sockets - even those to slower peers - to be acknowledged as well. That's completely FUBAR, but it's basically what fsync with data=ordered does to an entire filesystem. There are hundreds of performance…

Calling it obsolete is hyperbole. A standard definition of obsolete is something that was once common but no longer is. Just last week I had to revert a number of ext4 filesystems back to ext3 due to performance problems on RHEL5, so I'm afraid that ext3 will be in common use for years to come.

(The performance problems aren't present in Fedora 16, so you could argue that RHEL5 is obsolete.)

Re: Good news: Mozilla fixed their damn browser

#35

Post by Zirro » Wed, Dec 21, 2011, 2:51 AM UTC

otherpope wrote: post #32

kakuri wrote: post #25

Earlier quoted context omitted.

SRWare distributes a Chromium build called Iron that strips out privacy problems. http://www.srware.net/en/software_srware_iron.php

I used to recommend Iron but I don't anymore. I became increasingly uncomfortable with the lack of documentation and lack of transparency in the project. They delivered their source code as two code dumps on rapidshare last time I checked. Something just doesn't feel right at all.

http://chromium.hybridsource.org/the-iron-scam may be worth checking out, as well.

Re: Good news: Mozilla fixed their damn browser

#36

Post by PissedOffHNer » Wed, Dec 21, 2011, 3:33 AM UTC

Fixed?

I ran Google Chrome open with 80-100 tabs for 3 WEEKS without anything going wrong. Sunday the 18th I finally closed it and restarted my computer (after 3 weeks of non stop running).

Monday I launched the latest build of FireFox and within 30 minutes it crashed. It then crashed again a few hours later.

Goodbye FireFox.

Re: Good news: Mozilla fixed their damn browser

#37

Post by Athtar » Wed, Dec 21, 2011, 4:23 AM UTC

PissedOffHNer wrote: post #36

Fixed? I ran Google Chrome open with 80-100 tabs for 3 WEEKS without anything going wrong. Sunday the 18th I finally closed it and restarted my computer (after 3 weeks of non stop running). Monday I launched the latest build of FireFox and within 30 minutes it crashed. It then crashed again a few hours later. Goodbye FireFox.

This is unfortunately my main issue with FF. I am type of person that has multiple tabs open in multiple windows for days at a time and FF tends to choke in that scenario after a few days. I was hoping the 64-bit version might fix the issue but it's even worse.

It's still my main browser but this one issue just kills the entire experience.

Re: Good news: Mozilla fixed their damn browser

#38

Post by zobzu » Wed, Dec 21, 2011, 4:46 AM UTC

noelwelsh wrote: post #21

I've been running Aurora for a few months. It's been stable and gets the new toys quicker. Some add-ons break on some updates, but the few web dev. add-ons I use are always quick to update. I get questionnaires from time to time, but they don't take long to complete and are a trivial way to contribute to Mozilla. In short, give Aurora a try. It works as my full-time browser.

Note that you can disable the questionnaires if you don't like them (you do, but some may not)

Re: Good news: Mozilla fixed their damn browser

#39

Post by zobzu » Wed, Dec 21, 2011, 4:48 AM UTC

PissedOffHNer wrote: post #36

Fixed? I ran Google Chrome open with 80-100 tabs for 3 WEEKS without anything going wrong. Sunday the 18th I finally closed it and restarted my computer (after 3 weeks of non stop running). Monday I launched the latest build of FireFox and within 30 minutes it crashed. It then crashed again a few hours later. Goodbye FireFox.

I ran Chrome for 4 years (yeah) and closed it today after this news. Started Firefox Aurora and it crashed at opening.

Damn. Oh and this is my 3rd post.

Re: Good news: Mozilla fixed their damn browser

#40

Post by CPlatypus » Wed, Dec 21, 2011, 5:01 AM UTC

sciurus wrote: post #34

CPlatypus wrote: post #15

Earlier quoted context omitted.

The problem with data=ordered is that it imposes a global order, allowing one I/O stream to stall others. Imagine if waiting for data to be acknowleged on one TCP socket meant waiting for all data sent earlier on all sockets - even those to slower peers - to be acknowledged as well. That's completely FUBAR, but it's basically what fsync with data=ordered does to an entire filesystem. There are hundreds of performance…

Calling it obsolete is hyperbole. A standard definition of obsolete is something that was once common but no longer is. Just last week I had to revert a number of ext4 filesystems back to ext3 due to performance problems on RHEL5, so I'm afraid that ext3 will be in common use for years to come. (The performance problems aren't present in Fedora 16, so you could argue that RHEL5 is obsolete.)

Yes, the RHEL5 kernel is obsolete in that it lacks the barrier support necessary to implement proper behavior with decent performance. The fact that obsolete ext3 runs better on also-obsolete 2.6.18 kernels doesn't really tell us much except that sometimes two pieces of software evolve together, and sometimes versions that were developed together work better than versions that were not.