Re: [v6ops] About Req for Comments - "Transition to IPv6"

[Fred Baker <[e-mail]>]

> On Mar 5, 2020, at 8:46 AM, Owen DeLong <[e-mail]> wrote:
> 
>> Thank you very much for the pointer.  In it I could spot
>> the following footnote: "[4] IPv6-Only refers to network environments in
>> which use of the IPv4 protocol has been eliminated.”
> 
> IPv6-Only refers to network environments which are not using IPv4. That could be a network where IPv4 has been eliminated (rare at this time, though Facebook is a significant example) or it could be a greenfield deployment where IPv4 was never deployed.

<sans hats>

This, I think, is where the v6ops discussion has travelled in circles forever. To me, I would drop the word "network" and say that it is a specified environment. If I say that a system is IPv6-only, it is not using IPv6 internally or externally. It might or might not have IPv4 code in it, but it's not using it. If I say that a LAN, building, etc is IPv6-only, it is not to say that it couldn't have IPv4 in it if a DHCP server started handing out such addresses, but that it *doesn't*. If it is an Autonomous System, it means that only IPv6 is in use in that domain.

As soon as we get into the "but it could" or "was it translated from something else" or whatever rathole, we're into a discussion of angels, pins, and heads of pins. It will not end, because it can't. Someone will forever come up with a "what if" scenario to stir the pot.

Speaking for myself, I would have the user of the term identify the environment s/he is talking about, and say that *within*that*domain* communication is entirely accomplished using IPv6.

And STOP.