Re: V6 still not supported (was Re: CC: s to Non List Members, (was Re: 202203080924.AYC Re: 202203071610.AYC Re: Making Use of 240/4, NetBlock))

[Grant Taylor via NANOG <nanog () nanog org>]

On 3/11/22 9:39 AM, Joe Maimon wrote:
> I am not really convinced that IPv4 can be 
> ignored/marginalized/obsoleted without penetration reaching over 90%, 
> globally.

I feel like that's an unfair characterization / summarization.

The VAST MAJORITY of the pro IPv6 discussions that I see are targeting 
parity between IPv4 and IPv6.  As such, there is absolutely no ignoring, 
no marginalizing, no obsoleting of IPv4 in those discussions.

The vast majority of the discussions that I've participated in have not 
been IPv4 exclusive or IPv6.  --  The breakdown tends to be three 
categories, exclusive IPv4 (old), dual IPv4 and IPv6 (current), and 
exclusive IPv6 (far Far FAR future).

As I see it, if we divide the three categories equally, 0-33% is IPv4 
only, 34-66% is dual IPv4 and IPv6, and 67-99% (can be) IPv6 only.  -- I 
fudged the numbers a %, to simplify the 1/3 fractional math.  --  As 
such, we have crossed over from the exclusive IPv4 (0-33%) into the dual 
IPv4 and IPv6 (34-66%).  We have a long way to go before even 
considering exclusive IPv6 (67% (or higher)).

I believe that talking about removing IPv4 in any capacity /now/ is a 
disservice to the larger conversation.

I have my doubts about getting back to a single protocol Internet (IPv6) 
in my lifetime, much less my career.

> And until that point, IPv6 is an optimization, not a requirement.

How long do you wait during the "optimization" window before actually 
deploying IPv6?  The 11th hour?  Why not start deploying IPv6 with new 
green field deployments at the 2nd hour?



--
Grant. . . .
unix || die

Attachment: smime.p7s
Description: S/MIME Cryptographic Signature