On Thu, Jun 22, 2023 at 09:45:30AM +0200, Petr Špaček wrote:
> > I am confused, because I thought an IP address *was* a DNS name.
>
> It is, but the implication works only in one direction.
>
> Here's my reasoning:
>
> - Text representation of an IP address is a syntactically valid text
> representation of DNS name. (Again, hostnames are subset of all DNS
> names, so it probably is not a valid hostname.)
I disagree. For X.509 purposes the "text representation" of an IP
address is NOT what goes into an IP address SAN (unlike the case with
DNS-ID which does carry the presentation form of DNS names).
Also, IPv6 addresses (distinguished in X.509 IP SANS from IPv4 addresses
only by the length of the raw octet string) don't even look particularly
like DNS names in their presentation form.
Is "::1" a DNS name? Yes, pedantically, any string with at most 63
octets per label and at most 255 bytes total in its unescaped wire form
can be construed to be a DNS name. But it is rather a stretch to say
that either of "192.0.2.1" or "::1" are semantically DNS names for
purposes of this document.
Which does mean that clients connecting to a known service at an
"address literal" rather a domain don't have a way to represent this
as an SRV-ID at present.
--
Viktor.
_______________________________________________
Uta mailing list
Uta@ietf.org
https://www.ietf.org/mailman/listinfo/uta
