RIPE NCC Executive Board election

William Herrin bill at
Wed May 13 18:42:12 UTC 2020

On Wed, May 13, 2020 at 10:57 AM Brielle <bruns at> wrote:
> There's a fundamental lack of understanding from people on why you can't
> just 'bolt on' more addresses to IPv4.  Data structures in memory,
> software APIs and syscalls, hex notations, subnetting, route
> determination and propigation with internet routing protocols like BGP,
> hardware CAM, among many other things.

Hi Brielle,

Someone said much as you did way back in 2007. It bugged me, this
defeatism that said there was no way IPv4 could have been
incrementally updated to support more addresses, that a greenfield
protocol was the only path forward. So I designed an upgrade factoring
in the need for pre- and post-upgrade stacks and networks to
interoperate over a period of years. It took all of 4 printed pages.

It's clear IPv6 is the path forward. It was clear in 2007. But don't
for a second believe that's because IPv4 could not have been upgraded
in place. That's a failure of imagination.

I now return you to the scheduled brawl between Guilmette and Cohen.

Bill Herrin

William Herrin
bill at

More information about the NANOG mailing list