Re: IPv6 prefix lengths - how long?

[Mark Smith <[e-mail]>]

On Sun, 9 Jun 2019 at 03:03, James R Cutler <[e-mail]> wrote:
>
<snip>
>
> I have difficulty considering the use of a 64 bit as a “waste" because it:
> 1. Has not been shown to create a resource shortage in the foreseeable future.
> 2. It is a “safe” size for a pseudo-random space as far as we know.
> 3. As Philip Homburg stated, we have no proof for a minimum size space.
>

Waste occurs when you get no useful value from something.

As a relevant example, Ethernet addresses are 48 bits, yet when
Ethernet was designed, there was a limit of 1024 nodes on a link. That
means only 10 bits were required to uniquely identify a node.

So what value have those other 38 bits provided in the past 39 years,
and has that been useful value? Or does everybody think those other 38
bits are a waste, because they're strictly not necessary for the
protocol to work?


> I consider the effort and confusion involved in modification away from 64 bit Station IDs to be a waste (SLAAC, anyone?).
>
> I would rather see the effort put into better solutions security problems, including name to address resolution and routing.
> --------------------------------------------------------------------
> IETF IPv6 working group mailing list
> [e-mail]
> Administrative Requests: https://www.ietf.org/mailman/listinfo/ipv6
> --------------------------------------------------------------------