On Mon, Feb 27, 2006 at 04:15:40PM +0000, Thomas E. Spanjaard wrote:
> 
> It's a real pain that working out an appropriate new API with all the 
> new ALTQ features as OpenBSD (and FreeBSD/DragonFlyBSD) ha(s)(ve), turns 
> out to be so hard. I've seen several people come and go working on a new 
> API, and quite frankly my expectations of finally having a finished spec 
> alone diminish by the day. Perhaps core@ should set a direction for it 
> and start kicking people off of their asses ;).

You do understand that the KAME project committed to fix ALTQ so that it
could be integrated with packet filters other than PF, and then didn't
bother to do so in the subsequent year and half before it dissolved itself?

It's very difficult for the NetBSD project when external entities more or
less claim "ownership" -- that is, request that we coordinate all changes
through them -- to major sections of code, which prevents us from really
making any major enhancements ourselves, and then drop the ball.  This
has happened to us a number of times now; one other glaring example was
with IPsec, from which mess we are still in the midst of recovering.  But
at the same time, we usually make an effort to cooperate in this way
because, to be frank, we couldn't possibly produce or maintain all the
code that large outside projects bring in by ourselves.

The right balance is hard to strike.

-- 
  Thor Lancelot Simon	                                     tls@rek.tjls.com

  "We cannot usually in social life pursue a single value or a single moral
   aim, untroubled by the need to compromise with others."      - H.L.A. Hart