This is (slightly modified) from a recent post that I made. I figured I'd make a journal entry, because it was added to the story way too late to actually get any replies.
Here's what I'd do, if I were Mozilla CEO:
1) Apologize for all the zany shit that's gone on since 4.0. Can't hurt. Might help. At least, in the short term, it would mollify those who've recently complained that the Mozilla devs are uncaring, egotistical bastards.
2) Make sure that people who didn't speak for Mozilla made it clear they were speaking personally, not for Mozilla. They can say whatever they want, but they just have to make people aware of the fact that it's not a official statement from Mozilla.
3) Set up a new project, called Firecheetah or something, where development happens at breakneck speed, UI is in flux, and new technologies are readily adopted. It would have a rolling release, possibly automatically updated. This would be recommended for power users, enthusiasts, and web developers. There would be two or three channels: alpha, beta, and standard. Alpha would be like nightlies, where new code is tested first. These might crash a lot. Since the whole project is beta quality work, I'm not convinced that a beta channel even makes all that much sense, but, hey, why not. If it turns out to be redundant, it can be removed. Beta code is supposed to work as intended, but will probably need some testing to validate this. And, finally, we have the standard channel, where users are constantly inundated with new code from the beta channel, once it seems stable. If it turns out to not be stable, then they'll deal with it, since this project is more about rapid development than stability... and the next bugfix will be delivered shortly, anyway. This project is for the people who hate stagnation. It might be codenamed "Ritalin". Heh.
3) Retain the Firefox project, but as a more enterprise-friendly, stable version. Major versions would be heavily beta-tested before release. This would be the version that "normal" users and enterprises are encouraged to use. This version would be supported for a long time, with promises that the API will not change between minor revisions. Major revisions would be kept to a minimal amount each year, preferably with a leisurely development rate. Features from Firecheetah (ugh, I hate that name, but I'm not very good at branding) would be backported, after they've been sufficiently tested. My vision is for Firefox to be very conservative about changing the UI, adding superfluous features, or having crazy API changes that exist merely because someone got bored. This would be the browser that you use when you want a Firefox 2.0 oe 3.0 style experience: lean, mean, and stable. Architectural changes would be readily accepted, as long as they don't cause bloat or change the API. This project is for the people who hate change for the sake of change.
I'm sure there's stuff that I'm overlooking, and I might be overestimating the ability of computer users to handle having two browser projects coming from Mozilla, but I think this is actually pretty workable.
In the end, there isn't a huge difference between having two separate projects (Firefox and.. ugh... Firecheetah) vs having different channels of one project, where one of those channels (alpha or aurora or whatever they fuck they call it) is in constant flux. However, my way leaves a place for people who just want a browser that just works. I'm aware that my ideas aren't revolutionary, unique, or perfect for every circumstance; I'm just throwing this out there, because I'm bored and haven't posted a journal entry in years.