svn commit: r195200 - in head/usr.sbin: . wake

[Ken Smith]

On Tue, 2009-06-30 at 16:40 -0600, M. Warner Losh wrote:
> I hate to be blunt, but this leaves a bad taste in everybody's mouth,
> even if at the end of the day you get your way.

Translation: I did it again.  I hate it when that happens...  :-(

Two things happened here that were a bit unfortunate, and Marc/Martin
(most likely unintentionally) tripped over it.

The first thing that left a bad taste in some peoples' mouth is this new
thing got approved past the point I had said many times was to be the
end of new things getting added.  But people complained the warnings I
had given were buried in messages that had too much other information in
them instead of short and to-the-point messages.  So deadlines for a few
things got extended in short and to-the-point messages yesterday and
this pending request fell into one of those categories.

The second thing that left a bad taste in some peoples' mouth is that
this apparently (strictly judging by the fallout discussions) didn't get
as much peer review as some people would have liked.  Normally some
level of discussion happens on some public mailing lists (not private
email among a few potentially interested parties).  And even after that
happens and the commit gets made there is some time for fallout
discussions to happen.  Depending on the results of those potentially
lengthy discussions it might wind up being backed out.  But because of
the stage of the release cycle we're in me having approved this can be
viewed as short-circuiting the normal public review because odds are it
will wind up staying despite some peoples' opinions due to the stage of
the release cycle we're in.

That second thing is one of the things that I think we're stuck with as
part of the release process, but I really need help from you folks on.
It's very much like the rant I made about the commit requests.  And like
I said above I fell for this before, during the 7.0 release cycle.
During that release cycle it resulted in a rant with the subject line
"I'm not Head Dictator In Charge".  People complain if we lock out *all*
new additions, even ones as relatively simple as a new command, too
early in the release cycle.  So we *consider* allowing new stuff that
truly shouldn't impact the stability of the pending release until right
around now in the release cycle.  But we need to be able to trust that
if you send in a commit request for something new like this that you've
already done the peer review type stuff.  RE approval for something like
this doesn't trump normal peer review.  The alternative is to flat out
lock out *all* new stuff no matter how seemingly simple it is earlier in
the release cycle just to avoid the problem that maybe someone will try
to sneak something through without sufficient peer review.

Thanks...

-- 
                                                Ken Smith
- From there to here, from here to      |       kensmith at cse.buffalo.edu
  there, funny things are everywhere.   |
                      - Theodore Geisel |

-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 195 bytes
Desc: This is a digitally signed message part
Url : http://lists.freebsd.org/pipermail/svn-src-all/attachments/20090701/a8c372b6/attachment.pgp