Re: [v6ops] reclassify 464XLAT as standard instead of info

[Erik Kline <[e-mail]>]

On 25 September 2017 at 18:57, Alexandre Petrescu
<[e-mail]> wrote:
> Erik,
>
> I would like to ask you clarification.

I'm sorry but there is more clarification asked for and indeed
required here than I have time for.  Forget I said anything.

>
> Le 25/09/2017 à 11:47, Erik Kline a écrit :
>>
>> When speaking from the perspective of a client OS
>
>
> What is a "client OS"?  I think OSs are both clients and servers.
>
>> I have frequently observed and understood the phrase "IPv6-only network"
>> to mean that the link to which the client is connected only provisions and
>> forwards IPv6 packets.
>
>
> I agree with this definition of "IPv6-only network".  I would stress to
> also mean "no IPv4" - be it encapsulation, translation, or DNS-x-y.
>
>> IPv4 /may/ be available to the client as a service on these "IPv6-only
>> networks" within this definition,
>
>
> Yes.  But not as a "service".  I never heard about "IP as a service"
> other than some people at IETF, and other than in the context of a
> particular operator.
>
> IP is not a service.
>
> Services are at the services layer.  APIs are services, there is remote
> access service, etc.
>
>> whether by 464xlat or VPN or by some other means.
>
>
> VPN and encapsulations are also called "virtual networks".
>
> Is the "IPv6-as-a-service" a "virtual network"?
>
> If so, we want "IPv6 as a real network".
>
>> But that never seems to change the implied meaning of "IPv6-only
>> network" in these contexts.
>
>
> Depends who you ask.
>
> Alex
>
>>
>> Just an observation.
>>
>