The guys and gals over at the Mozilla Corporation have been making remarkable strides on the next version of everyone’s favorite browser. Firefox 3 has been in beta testing for months now, running late from its initial target of October 2007. (I can’t find any citations for that, so just take my word for it.) Fortunately the delays are paying off. Take a look at a section of the release notes for Beta 3.
[Improved in Beta 3!] Memory usage: Over 350 individual memory leaks have been plugged, and a new XPCOM cycle collector completely eliminates many more. Developers are continuing to work on optimizing memory use (by releasing cached objects more quickly) and reducing fragmentation. Beta 3 includes more than 50 improvements to memory use over the previous beta.
That in itself is fantastically impressive. Maybe we won’t see so much Hungry Hungry Firefox RAM-gobbling in the near future.
While I acknowledge that such massive code cleanups require a considerable number of man hours, I have to ask myself whether that’s really enough to keep Firefox 3 running so late. My theory is that although the code changes might be significant, the real drain on time is the constant reworking of the Mozilla robot for each new beta release.
I mean, don’t you think that’s a bit much for each release? I applaud their focus on presentation, and I’m sure it pays off to treat your beta testers well, but really now. You guys can’t think of anything better to do with all those ten of millions of dollars of Google bucks you’ve been getting annually? The bubble burst, my ass.
At least the newest beta (Beta 5 at the time of this writing) is rock solid and fast as hell.