My proposed compromise at the trans workgroup meeting yesterday was to
change the MUST in the following paragraph to a SHOULD:
"Logs MUST accept certificates and precertificates that are fully valid
according to RFC 5280 verification rules and are submitted with such a
chain."

I believe that this change allows creating logs that can reject valid
submissions under some circumstances while complying with 6962-bis. And we
wouldn't have to go through WGLC again for such a change.

On Sat, Nov 5, 2016 at 12:42 AM, Ben Laurie <[email protected]> wrote:

> On 4 November 2016 at 14:07, Peter Bowen <[email protected]> wrote:
> > On Fri, Nov 4, 2016 at 6:33 AM, Ben Laurie <[email protected]> wrote:
> >>
> >> The bit you didn't quote does say they the log has to accept valid
> >> certs, tho: "Logs MUST accept certificates and precertificates that
> >> are fully valid according to RFC 5280 [RFC5280] verification rules and
> >> are submitted with such a chain."
> >
> > Sorry about that.  So the three MUSTs together are:
> >
> > - Logs MUST verify that each submitted certificate or precertificate
> > has a valid signature chain to an accepted trust anchor, using the
> > chain of intermediate CA certificates provided by the submitter.
> >
> > - Logs MUST reject submissions without a valid signature chain to an
> > accepted trust anchor.
> >
> > - Logs MUST accept certificates and precertificates that are fully
> > valid according to RFC 5280 verification rules and are submitted with
> > such a chain.
> >
> > When I read these together, I read that Logs must accept _any_
> > certificate that is fully valid according to RFC 5280 verification
> > rules and chains to any root the log trusts and logs must _only_ log
> > such certificates (and no others).
> >
> > If this is accurate, we need to account for all types of certificates
> > being logged, as a log cannot choose to reject certificates for usages
> > other than server authentication and the log cannot reject
> > certificates that have personal information (e.g. an server
> > authentication certificate that states which human requested the
> > certificate in the subject).
> >
> > This seems like a very strong assertion of policy rather than a
> > technical discussion of how the CT protocol works.  I would again ask
> > the WG to reconsider the requirement levels specified in this section.
>
> FWIW, I tend to agree. I would also note that operationally, we have
> already discussed options for logs that are incompatible with these
> requirements, for example:
>
> * sharded logs (e.g. by hash(cert) mod n).
>
> * logs that only accept short or long lifetime certificates (this
> would allow the working set to be reduced in size for CAs that issue a
> lot of short-lived certs).
>
> _______________________________________________
> Trans mailing list
> [email protected]
> https://www.ietf.org/mailman/listinfo/trans
>
_______________________________________________
Trans mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/trans

Reply via email to