Esko Dijk <[email protected]> wrote:
    > The current BRSKI text to me suggests that CA:TRUE is a requirement for
    > the pinned-domain-cert. But I'm okay with not having CA:TRUE for this
    > certificate, as you propose, in which case I think the BRSKI text needs
    > some minor updates on the wording.

    > For example, if it's just a Registrar cert with CA:FALSE and RA:TRUE
    > then it shouldn't be called a "domain CA" cert or "domain cert".

    > If the Registrar is not a CA, it does need to be a Registration
    > Authority (RA). (See Section 2.5.3 / 2.5.5 / 5.5.4 /
    > https://tools.ietf.org/html/rfc6402#section-2.10 )

    > So the requirement for the pinned cert is that it is either RA or CA.
    > (Both seems also possible to encode in the cert, although that seems
    > equivalent to a CA.)

I think you that you are likely correct: every entry in a chain which is not
the EE, has CA:TRUE, and the last one, is the Registrar which ought to have
cmcRA set.

If I were to locate a PKIX-lawyer hat, I would say that the Pledge SHOULD
check this.   Given that there is a strong desire to operate registrar's
using only certificates obtained via ACME (LE), and it will be difficult to
get cmcRA set... I don't know what to do here.
  <Please insert long EMU Thread here>

We have written in the ACP document that a device SHOULD verify the cmcRA bit
is set on any EST server that the device speaks to to renew it's certificate.

Please note that I am saying "device" and not pledge!
We did not insist on cmcRA be set in section 5.1, specifically because we
didn't know if RA:TRUE was something we could insist upon here.

--
Michael Richardson <[email protected]>, Sandelman Software Works
 -= IPv6 IoT consulting =-



Attachment: signature.asc
Description: PGP signature

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

Reply via email to