Re: [IPv6] [Last-Call] Artart last call review of draft-ietf-6man-rfc6874bis-02

On Mon, Mar 27, 2023 at 1:39 PM Brian E Carpenter <
[email protected]> wrote:

> On 28-Mar-23 09:19, Rob Sayre wrote:
> > Hi, firstly, this is a subtle and unpleasant problem.
>
> Fully agreed. I should have responded to another of Roy's points, however.
> Unfortunately, the problem is not changing RFC 4007 to use a better
> delimiter. The problem is changing every IPv6 socket library implementation
> and an unknowable number of application programs, help files, manuals,
> books, and brains, that have embedded the current delimiter. That's roughly
> 18 years of running code, and it would take us another 18 years at least.
>

Here, we don't agree. The first reason is that URIs are almost never
straight copy/paste operations. IP addresses are the small problems. :)

You could instead influence the representation, as IDNA does.

I also still agree with Roy on principle, but when I read the Java 8 URI
documentation, I gave up. It's out there and I can't influence it. I think
the draft should be more direct, as I said. The security problems are also
baked into the grammar, as accurately described, so you really do need to
cover that part too.

thanks,
Rob

_______________________________________________
art mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/art