Proposed technical roadmap - addendum

[Eric S. Raymond]

I left out someone...

Mark Atwood:
   1. Write up the release procedure in devel/hacking.txt
   2. Decide timing of our betas and 1.0.

Even leaving out that release scheduling is entangled with the
external politics that is Mark's PM job, I don't want to have to
think or worry about that end myself while I'm trying to get
TESTFRAME working.

Of course Mark Atwood wearing his PM hat has to approve anything
I suggest about allocating Mark Atwood the dev's time... :-)

My scope estimate for TESTFRAME is roughly 90 days.  However, this is
one of the unusual cases in which things could break so it lands
*faster* than I expect.  The capture side is relatively simple
and easy to estimate; the replay side is where the uncertainty is,
and it's uncertain in both directions.

My scope estimate for constructing a replay branch is, as previously
noted, 4-5 (grueling) days.  I'm prepared for this to interrupt
TESTFRAME, but ideally nothing else should.
-- 
		<a href="http://www.catb.org/~esr/">Eric S. Raymond</a>