North American Network Operators Group

Date Prev | Date Next | Date Index | Thread Index | Author Index | Historical

RE: OMB: IPv6 by June 2008

  • From: Scott Morris
  • Date: Thu Jun 30 17:37:04 2005

Heheheh...  But see, wasn't that one of the whole theories behind the
"aggregation" schemes built into the allocation of IPv6 address?  Come
now...

Because we have deployed it today in a manner where that's not possible
doesn't make it a "rule" per se.

Is this theory any different that simply filtering the multiple allocations
denoted as RIPE or APNIC allocated IPv6 chunks?  I'd think not.

*shrug*  You're reading way too many politics into this, but not seeing the
designs of IPv6 in the same light.  SSDP.  (Same .... Different Protocol) 

Scott

-----Original Message-----
From: Andre Oppermann [mailto:[email protected]] 
Sent: Thursday, June 30, 2005 5:27 PM
To: [email protected]
Cc: 'Fergie (Paul Ferguson)'; [email protected]; [email protected]
Subject: Re: OMB: IPv6 by June 2008

Scott Morris wrote:
> We could have been much better served adding 3-bits at the beginning.
> Effectively giving a full IP v4 space to every continent (even 
> Antartica) and having an extra one for the extra-terrestrial working 
> group.  ;)
> 
> And it would have given us real geographic-based filtering 
> capabilities at the same time without any major changes to everything 
> we have worked so hard to get to the level of insanity where we are today.
> 
> *shrug*  Simple things often get overlooked.

bzzzt...  You just described a rule #1 violation; IP addresses are routable
entities and thus by definition unsuitable for any kind of geo-location.

Rule #2 would be that IP addresses do (and must) not encode routing
information, they just serve to transport data.  All routing information is
carried on the routing layer and applied to the forwarding layer from there.

When do people learn that these layers do not intermix just like water and
oil do not?  I guess the only lession history teaches us is that it doesn't.

--
Andre