On 8/23/22 17:43, Paul Hoffman wrote:
On Aug 23, 2022, at 2:00 PM, Joe Abley <jab...@hopcount.ca> wrote:
I may have missed something (I have been trying very hard) but it seems a
little weird for the wire format for a definitively non-existent domain name to
be specified at all, to be honest; I'm not sure what imagined audience that is
intended to help.
This is a very good point, one that I had not considered. Applications that
know about the non-DNS will be getting input both from users or other
applications, and thus get their input case-mixed. The same will be true for
recursive resolvers.
I'm not sure why you conclude that input would be case-mixed.
Unaware applications: yes, perhaps mixed; but as they're unaware, they'll
ignore the carve-out regardless of case
Aware applications: ... will produce only what's compliant. And the question
here is what we want to define as compliant with the carve-out.
I see no problem with specifying one (1) lowercase variant only. All non-DNS
would be invited to reside underneath. Am I missing something?
If this thinking is correct, the addition of "corresponding to a 0x03616c7400 suffix
in DNS wire format" was wrong. However, I'm not sure what we want to say instead
that will be clear enough for developers to act on.
We should decide what the set of carved-out binary TLD strings is, and then
enumerate them for maximum clarity to DNS implementors who want to have an
answer about what exactly to (not) forward upstream etc.
Best,
Peter
--
https://desec.io/
_______________________________________________
DNSOP mailing list
DNSOP@ietf.org
https://www.ietf.org/mailman/listinfo/dnsop
