>We'd like others to review and encourage further discussion relating to these 
>drafts.  Thank you for your time.

I'm quite uncomfortable with the bit that says you look up the policy
at https://policy._mta_sts.example.com/current

By my reading of RFC 3986 which defines URIs, that's not a URI.
Section 3.2.2. does have ABNF which allows underscores in the host
part, but it then has text on page 21 that says that names looked up
in the DNS are RFC 1123 hostnames that are letter, digit, hyphen
only.  It says the underscores are intended for names looked up in
other ways.

My CA (Startcom) won't sign names with underscores, and I doubt that
other CAs will, which makes them rather unappealing.

I have two suggestions, a reasonable one and a gross kludge.

The reasonable one is to use SRV, so it'd be like this:

 _sts._tcp SRV 0 0 443 sts-policy.example.com.

Bad guys could spoof your SRV record without DNSSEC, but they can also
spoof the A record for policy._mta_sts.example.com so that horse left
the barn.  You could require that the target of the SRV has to be a
subdomain of the original domain which, along with the signed SSL
certificate, would limit the scope of evil.  This proposal has the
advantage of being built on top of well established standards, and the
disadvantage that looking up SRV records is still alleged to be hard
in some contexts (but I don't think ones that matter here.)

The gross kludge is to use xn--sts0 as the tag, e.g.

 https://policy.xn--sts0.example.com/current

The string "sts0" is deliberately invalid punycode, so while xn--sts0
is a valid hostname label, it's not an A-label and will never appear
in an IDN hostname, Hence it's very unlikely to collide with other
uses.  I hope I don't have to explain why it's a kludge.  My CA will
sign it. I checked.

R's,
John

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

Reply via email to