Re: commit access policies

[Alfred M. Szmidt]

   If the decision is that the code in gnumach-1-branch is what we
   should care most about using and releasing, and that this is the
   code which should be the basis of ongoing development, then it is,
   de facto, HEAD.  Which is what you said.

Right.  And if the decision is that, then the bug-fix-only rule _is_
lifted for the code in gnumach-1-branch.

   In that case, it should be incorporated into HEAD and the existing
   HEAD be a branch, with a suitable name.

I don't see the whole fuss about adding features to gnumach-1-branch,
oskit-branch got features added while gnumach-1-branch was HEAD.  The
only really good reason for swapping branches is that the code in
right now HEAD is b0rked, not that gnumach-1-branch is the most cared
about.

In either case, I really don't care if there is swapping branches, or
that someone has to jump on one leg for a week, aslong as one can add
featires to the code that is currently known as gnumach-1-branch.

   If this is not immediately crystal clear, then you shouldn't be
   making these decisions.

It was crystal clear from the beginning, you simply didn't understand
me.