Re: The bottom is /112 (was: RE: Extending a /64)

[Mark Smith <[e-mail]>]

On Wed, 18 Nov 2020, 07:50 Joe Maimon, <[e-mail]> wrote:

>
>
> Mark Smith wrote:
> >
> >
> > On Wed, 18 Nov 2020, 03:40 Joe Maimon, <[e-mail]
> > <mailto:[e-mail]>> wrote:
> >
> >     Class E syndrome
> >
> >
> > We can deal with that if it happens.
>
> If it does happen, it will happen 2-5 times longer than it did for ipv4,
> and it will be all the worse then.
>
> The only historical example we have suggests that the rest of the first
> 3 bits cannot be fully counted on.
>

You're simplifying the situation and pretending there wasn't a better
alternative available.

IPv6 was available and ready to be deployed when the idea of using Class E
came up.

The cost of upgrading networks and hosts to use Class E would be about the
same as deploying IPv6.

Which provided the better return on time and resource investment?

Spending it on Class E wouldn't have avoided the IPv6 deployment cost, it
would only have delayed it, and effectively doubled the cost of moving to a
larger Internet address space.


> > It wasn't worth dealing with that in IPv4 because it didn't give back
> > enough address space for the effort.
> >
> >
> Thank you for helping define class E syndrome. Which is a symptom of
> community standardization mistaking themselves for a top down hierarchy
> and creating self fulfilling obstacles to everyone else.
>

IPv4 addressing history and motivation is a lot more complicated than that.

As an example, have a look at the IPv4 address format in RFC760, and then
compare that RFC's date to RFC791's, where address classes were introduced.

 Also look up the address format in RFC675, effectively IPv1 (before TCP
was split away).

IPv6 is really the first protocol designed for the Internet that we have
today. IPv4 was an experiment that escaped the lab into world wide
production.


>
>
>