On 8/23/22 11:38, Paul Hoffman wrote:
On Aug 23, 2022, at 7:47 AM, Warren Kumari <war...@kumari.net> wrote:
So, I'd think something like: "For compatibility with existing applications and to
maximize interoperability, it is recommended that names that end in .alt follow DNS name
syntax." (or something similar but better worded).
An application that understands a non-DNS protocol already is compatible with
that naming scheme.
I think their point is that the application (e.g. browser) may be agnostic of
the resolution system (= accept the name), but resolution may fail because
something switching layer like nsswitch would choke on a non-DNS-style name,
*even when* the downstream non-DNS resolver would be available.
So, by making the non-DNS names DNS-style, one can allow for more agnostic
intermediate layers in the process that make DNS-like assumptions.
Given that, I can see why a sentence like the one suggested by Warren could
make sense.
OTOH, it's possible that a given non-DNS protocol comes with its own
application layer (i.e. no legacy pieces with DNS-like assumptions involved),
and it's questionable whether a name syntax recommendation would cause undue
restrictions for these cases.
All in all, I think either way is fine (weak suggestion/recommendation, or
nothing at all).
Peter
--
https://desec.io/
_______________________________________________
DNSOP mailing list
DNSOP@ietf.org
https://www.ietf.org/mailman/listinfo/dnsop