Fast as my browser

By Percy Cabello

Firefox 3.5 will be fast. Twice as fast as Firefox 3 to be precise. The question is, are you in the same league? Not rendering pages or running web applications (now, that would be seriously cool), but in any other skill. That’s what Fastest Firefox is all about.

The latest Mozilla marketing project is asking you to submit a video of yourself performing some task really really fast.

Fastest Firefox image

To ignite the fire, videos of actual world record holders are being posted (starting with the fastest clapper) at the site. Just ensure not to use any copyrighted material, keep it under 30 secs/100 MB, encode as MP4, WMV or QuickTime (yes no Theora, but videos are converted to this format anyway) and submit it for others to enjoy before June 21.

You could win a Firefox 3.5 T-shirt if your video is selected for a video compilation.

Posted on June 2, 2009 - 12:00 am || More on Contest, Events, Firefox 3.5

Comments

Niels R.

June 2, 2009 12:00 am

Does this mean that Firefox 3.5 will be loading faster on both Windows and Linux when you start it the first time?

Or are you only talking about rendering times?

Reply

James

June 12, 2009 12:00 am

According to my SunSpider tests on a Windows XP machine, Firefox 3.5 beta 4 is about 2.4 times as fast as Firefox 3.0.x (x is 10 or 11) , and the result is pretty stable from test to test. Yesterday I did another bunch of SunSpider tests on another machine, which dual-boots Windows XP and Fedora 11 (two OS’s on same hardware). Firefox 3.0.11 runs on Windows XP and Firefox 3.5 beta 4 runs on Fedora 11. Surprisingly, beta 4 is about 50% to 100% slower in Linux (Fedora 11) than the 3.0.11 in XP, and the Linux result is quite unstable from test to test, with most of the results fall in the range of “50% to 100% slower” and some slower than 100%, although I had tried to make sure the only app running in Linux is Firefox.

A Google search told me that this was not a new discovery. Firefox in Linux has been notoriously slow for quite a while and, what’s more, it seems nobody is sure about the true cause. The closest I can come up is a performance issue on Linux kernel fsync() call, which is kind of fixed in kernel 2.6.30 (Fedora 11 uses kernel 2.6.29). But I don’t know if SunSpider script will cause disk I/O in Linux (and thus can perhaps touch fsync()). Even if it does, it might not account for 100% of the Firefox slowness…

Reply

Leave Comment