Re: commit access policies

[Alfred M. Szmidt]

   http://lists.gnu.org/archive/html/hurd-devel/2002-05/msg00060.html

   Here, Roland clearly states that gnumach-1-branch is only for
   bugfixes.

That was 3 years ago, when OSKit was all the rage.  gnumach-1-branch
is far far simpler to maintain, as has been shown by the continues
efforts to keep it working as it should including bug fixes, new
drivers, and what not.  I don't even think that HEAD actually works at
all right now, and I really don't care.  The benefits of having
everything in one tree outweigh the benefits of spliting it up just
cause there are more drivers in OSKit that nobody actually uses.  Most
NIC's are now supported by GNU Mach 1.x + my NIC patch.  This and the
SCSI drivers are really the only drivers that we even remotley
support.  The change to OSKit doesn't add anything, other than a
headache.

So there is no reason not to revert the `bug fixes only' rule for
gnumach-1-branch, if there is a tier one who checks patches.