The roads I take...

Here's a very very late summary of SeaMonkey/Mozilla-related work I've done in week 30/2008 (July 21 - 27, 2008):

• SeaMonkey & Thunderbird on Mercurial:
  The switch to Mercurial was the big story of that week, including the new comm-central build system. Everything has worked fine, SeaMonkey's build machines were updated smoothly, unit test machines are a bit bumpier and still see test failures due to new code in 1.9.1 that we didn't have on cvs trunk before.
  I also reviewed a couple of followups on client.py and did a few related changes after the switch myself.
• SeaMonkey on Maemo:
  This was just a fun thing to try out and is not a project we have plans to follow up with, but I've done a Maemo build an could actually run SeaMonkey on my N810! Even if there are some slight problems around there, I find SeaMonkey's ChatZilla more usable on it than the Pidgin IRC client adapted for that platform.
• L10n:
  As mentioned before, lost L10n builds on SeaMonkey trunk for the time being with the switch to Mercurial. We're planning to reinstate them as soon as the new processes for 1.9.1 are in place and we can adopt them.
• Various Discussions:
  Mozilla Summit, Mercurial move, etc.

Sorry for this arriving that late, but I didn't get around to doing it when I was preparing for or at the Mozilla Summit, and of course not in vacation.
The Summit itself was a very successful event, mainly of connecting with people and meeting people face-to-face that I only knew the names of or had email/IRC/bugzilla contact before. The sessions were also interesting, though there were too many things I would have wanted to be at in parallel. Oh, and we had a very fruitful, long and late breakout session about release automation on Mercurial, the processes that will be set up will be made in a way that they can work with Thunderbird, SeaMonkey and Sunbird from the beginning. Thanks to the MoCo build & release team for including us there and making this possible.
I'm still digging through lots of backlog from those two weeks, so expect it to take some time still until I'm back to normal reaction times on whatever issues are out there.
And yes, the first alpha is definitely nearing, target is early to mid September now, nearly in sync with Thunderbird 3 beta 1.