RSSThere's now a Firefox 3.5 Release Candidate 3
By Scott M. Fulton, III | Published June 25, 2009, 1:58 PM
Early Betanews tests indicate some noticeable changes in the performance profile of a Release Candidate for Mozilla Firefox 3.5 that was posted to the organization's servers this morning. Overall, performance has improved by 1.7% in RC3 over RC2, released Monday, although underneath the surface, there appear to be some big gains and some setbacks.
Regular string expression (RegEx) handling appears to have suffered in the latest release, with SunSpider performance benchmarks indicating RC3 required 100.4 ms to complete that heat, versus 85 ms for RC2. Other SunSpider scores were marginally slower, though RC3's scores in rendering and the Celtic Kane battery helped RC3 to more than compensate. Array object handling, for instance, improved by an astonishing 40% in RC3.
We also noticed what appeared to be a JavaScript event processing error in RC3 in Win7, in Betanews tests this afternoon, although we have not yet confirmed whether the browser itself is to blame or whether the event is a fluke.
In all, Firefox 3.5 RC3 in Windows 7 posted a Betanews index score of 9.0 versus 8.85 for RC2. In other words, RC3 is now performing at nine times the level of Microsoft Internet Explorer 7 in Windows Vista SP2. The latest beta channel release of Google Chrome 3 also made its way to Fileforum today, following up on the latest Opera 10 Beta release a few days ago. So soon we'll reset the field for you, and give you another full-field rundown.
I wish they'd simply stop issuing new RCs and wait for the final release. My experience with the RCs has been universally bad. It is not, I repeat, NOT the user's responsibility to fix their browser. They're certainly not going to pay you if one of these browsers crashes your system.
I don't rely on any browser to secure my system. That's why I use a hardware firewall and ESET's System Secure.
Score: 0
|Anyone for IE8 ?
Score: -1
|lol those speed tests must of pissed off firefox
Score: 0
|Does a browser have emotions?
Score: 0
|ya you didnt know
Score: 0
|The changes in the nightly builds have been so slight that they barely spend any time being downloaded. It looks ready, so I'm just hoping that the extensions are working shortly.
Currently, with only Betanews open, real memory is at 93.88 MB, which seems better than Firefox 3.0.11, but still not very good. The browser seems effortless though, and it's been a long time since browsers just seemed to glide along.
Score: 0
|Hope the RC 2 bugs got fixed.
Score: 0
|You can download an Intel-Optimized version of Firefox 3.5 RC3 at http://www.latko.org/downloads/
Score: -1
|"You can download an trojanized version of Firefox 3.5 RC3 at..."
fixed that for you
Score: 0
|What do you care, fatty? Trojan's don't work on Mac OSX, right?
Score: 0
|You've made full news stories on all 3 of the RCs. Do you really not have anything better to talk about? 1.7% change? ZOMG! I guarantee you will never notice that in actual use. And in any case, if there is another RC, these results become meaningless. No reasonable person is going to continue running the RC for long after a final is released, even if there is a small performance drop. Best just to see these numbers for the final release and be done with it. Not that I begrudge one or two previous articles talking about the promised performance of 3.5, mind you. But at this point it's just ridiculous.
Score: 0
|In Scott's defense, this is Betanews...
Score: 1
|Scott, mind sending me or posting the error message you're talking about? We're not seeing any such event processing errors in Javascript.
Also, I suspect that the changes you're seeing are more related to random noise in the benchmark, as we really didn't make that many changes to the JavaScript engine in RC3.
Score: 0
|1.7%
Surely this benchmarking is boring you by now?
Score: 0
|What benchmarking? I'm just here for the witty repartee. ;)
I don't think anyone's been here for the benchmarking since the whole VM/bare metal issue. ;p
Score: 0
|