Some thoughts on this:

A 3 0 0 tlsa will work as well as a 3 1 x.  The client can pull the spki
out on its own.

I find both possibilities,  incorporation in the tla rfc or a new rfc
from us equally accecptable methods of publication.  As long as our
input is reflected.

It is not that *all* tlsas need to be limited just because the server
supports the oob methods.  Rather, *at least one* of the published tlsas
must be usable (3-0-0 or 3-1-x and secure in this case) and match.

Should a different kind of bare key show up, one which doesn't look like
the spki blob expected by x-1-y tlsa (openssh keys would be interesting),
we can publish an rfc defining matching type 2.  But given:

,----< excerpt from draft-ietf-tls-oob-pubkey-11.txt >
| namely, the SubjectPublicKeyInfo structure of a PKIX certificates
| that carries the parameters necessary to describe the public key.
`----

we do not need match 2 for this.

-JimC
--
James Cloos <[email protected]>         OpenPGP: 0x997A9F17ED7DAEA6

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

Reply via email to