44/8

Mon Jul 22 22:17:02 UTC 2019

The fundamental reason given, from several sources, was that our experience with IPv4 address trading says that no matter how many IPv4 addresses we create or recover, we won't obviate the need for a replacement protocol. The reasons for that are two: (1) IPv4 isn't forward compatible with anything (if it had a TLV or equivalent for the address, we could have simply extended the address), and (2) 2^32 is a finite number less than the number of addressable entities in the world. Yes, it would be interesting to use Class E as unicast space. The instant we make it possible, it will be bought up by companies and countries desperate to delay their IPv6 deployment - and we will then, once again, be out of IPv4 space.

We even had a guy write five internet drafts about how it is possible to enumerate more than 2^n entities with an n bit number.

Speaking for myself, I don't see the point. It doesn't solve anything, and I'm not sure it even meaningfully delays anything. The time has come to move to a protocol that allows us to enumerate the set of addressable objects without losing our minds.

> On Jul 22, 2019, at 3:04 PM, William Herrin <bill at herrin.us> wrote:
> 
> On Mon, Jul 22, 2019 at 11:56 AM andrew.brant via NANOG <nanog at nanog.org> wrote:
> Whatever happened to the entire class E block? I know it's reserved for future use, but sounds like that future is now given that we've exhausted all existing allocations.
> 
> The IPv6 loonies killed all IETF proposals to convert it to unicast space. It remains reserved/unusable.
> 
> Regards,
> Bill Herrin
> 
> 
> -- 
> William Herrin
> bill at herrin.us
> https://bill.herrin.us/