Re: draft-bourbaki-6man-classless-ipv6-00

[Lorenzo Colitti <[e-mail]>]

On Fri, Jun 16, 2017 at 10:36 AM, Brian E Carpenter <
[e-mail]> wrote:

> I think that portability out of the box is a much stronger argument. Peter
> has chosen to hand-craft his networks.


... and hand-crafting networks is precisely what network operators do. It
is, after all, their job.


> Most people (99.99%) simply unwrap
> their new device, ignore the "quick start" guide, and switch it on. There
> is an enormous market incentive for that to work 100% of the time, which is
> an enormous incentive to stick with /64.
>

Your premise is correct but your conclusion is wrong. Yes, there is an
enormous incentive for things to work 100% of the time. But that does not
mean stick to /64. Quite the opposite - it means "work regardless of the
prefix length on the network". Before you disagree with me, observe the
fact that most implementations work on Peter's network, today, even though
Peter's network is in complete violation of RFC 4291.

Not to mention the late lamented /48 recommendation to ISPs. Again,
> nobody is trying to sabotage /64 as the norm for SLAAC networks, and
> nobody is trying to sabotage SLAAC as the norm for plug and play.
>

I know that you're not *trying* to do that. I'm trying to make you realize
that by pushing for a non-fixed boundary you *are* doing it.

As for "nobody" - it's not true that nobody is trying to do it. If you read
Job and Peter's statements carefully, you'll see that they are.


> I agree. But at the same time we need to be honest in our specs. RFC4291
> does not describe the deployed architecture accurately, which is why
> we have to fix it if we're serious about Internet Standard status.
>

That's not true for any value of "accurate" that is meaningful in the real
world. I'd argue that RFC 4291 (with the exception made in 6164), is
accurate for at least 99.9% of links on the public Internet.

Making it accurate for the remaining 0.1% is not worth removing the text in
the standard that stops operators like Peter being able to claim that hosts
should support addressing models that are bad for their users.