DragonFly future planning

I've got some ideas for how we're going to have the next release, but I
wanted to write them out publicly.  There's a bunch of changes going on for
DragonFly, and it'll be neat to have them in the next release.  If I name
you in this writeup, tell me if I'm overcommitting your time.

- 3.4 was released in April.  We aim for 6-month cycles, which means 3.6
should be near the end of October.  That gives us about a month and a half
to deal with integrating GSoC 2013 code.  If you are a mentor or student,
tell me if that's possible/not possible.

- dports is changing to gcc 4.7, as John Marino posted.  We'll probably be
using dports built with gcc 4.7 for DragonFly 3.6 release.

- We have i915 support, thanks to several people's hard work.  I think it's
OK to plan on having it as part of 3.6.

- Matt's working on Hammer 2.  It's not going to be done for 3.6, but it
would be neat to have an experimental version for people to try.  I know
Matt's got his hands full, but if someone wanted to work on Hammer 2
reclaiming deleted space, it would probably be at least easier to
experiment with.

- 3.6 will probably be a dports-default release, because dports has pretty
much swept the board with people adopting it over pkgsrc.  The installer
for DragonFly assumes pkgsrc right now, but I think Sascha has been working
on a conversion. So, start thinking about a pkgsrc -> dports conversion if
you haven't changed over yet.  (pkgsrc will still work on DragonFly for
those wanting to go to pkgsrc-2013Q3, for instance; just I don't think
there's anyone actively working on it for DragonFly now.)

Looking farther ahead, with some things that I certainly won't be doing but
would love to see:

- As John Marino wrote recently, clang will probably need to come in since
FreeBSD's moving to it.  I'm happy to let them go first, especially with
ports.

- Replace groff with mandoc?  This idea has been broached before.  I think
groff is the last c++ item to clean up.

- After 3.6, if Hammer 2 gets to a point where you can boot from it and
generally run normally, I'd suggest the next DragonFly release after that
is numbered 4.0.

- A new DragonFly GUI image - this would be part of switching to a dports
install.

- Evaluating i386 platform support.  I don't think it's time to drop it
yet, but it's worth thinking about again within 6 months time.

Please add to my list if there's features you're planning on working on for
this release/next release, or if I'm very wrong on something.