IPv4 and ASN Policy draft on-line

Hi,

On Tue, Sep 25, 2001 at 12:58:44PM +0200, Wilfried Woeber, UniVie/ACOnet wrote:
>   Wrong.
>   there are valid cases where a (globally unique) AS# is required which
>   does not show up in the DFZ.  (and, of course, there are places where a
>   private AS# would do just as well).

Hmmm.  I'm curious: could you name a few?  The things that I could
imagine are "private peerings somewhere", which "should" work using
private AS#s just fine.

>   The bottom line is: how much are we (collectively) prepared to pay (in
>   real money for RIR staff salaries, complexity, hassle and delay for
>   *all* requests) in order to save a couple of AS numbers...

Or to phrase it differently: will it be more expensive than to upgrade all
routers world-wide to be able to handle 32bit AS#s?

While there *is* the option to never reclaim AS#s and just go to 32bits,
I, for one, do not really want to handle 32-bit-long AS#s in daily
operation, and do not really want to see all the new and exciting bugs
that they would bring with them...

Gert Doering
        -- NetMaster
-- 
SpaceNet AG                 Mail: netmaster at Space.Net
Joseph-Dollinger-Bogen 14   Tel : +49-89-32356-0
80807 Muenchen              Fax : +49-89-32356-299