Re: [therightkey] review of draft-laurie-pki-sunlight-03

[Stephen Farrell <[e-mail]>]

Just the bits where there's something to say:

On 12/19/2012 12:17 PM, Ben Laurie wrote:
> On 17 December 2012 01:54, Stephen Farrell <[e-mail]> wrote:
> 
>> - Where're the OIDs in 3.1 and 3.2 registered? They should be
>> in the some registry I guess.
>>
>> - 3.1, shameless self-promotion - draft-farrell-decade-ni has
>> a way to hash public keys:-) Yes, filling in the TODO is
>> needed.
> 
> Am I allowed to refer to I-Ds?

You could. However, my I-D is gonna be stuck for a while
since it has a normative dependency on http/1.1 drafts so
unfortunately you're probably better to just copy the bit
of text you need from there, if you want to use it. If
you do that you could copy what you need and add an
informative reference and it could all be fixed up later
without slowing down your RFC.

Or you could refer to DANE [rfc6698] assuming you hash
the SPKI which both my draft and DANE do.

>> - intro, para 2: I'd avoid the word "prove" since mistakes can
>> be made (e.g. a clock might be off)
> 
> I am slightly at a loss for a better word here!

s/prove/provide strong evidence /  maybe

>> - 3.1, why "0..." for X509ChainEntry and "1.." for
>> PrecertChainEntry? You probably ought say and that might need
>> to change if you change the "same CA issued both EE cert and
>> precert-signing-cert" rule.
> 
> I can't parse this!

Can't say I blame you, wonder who typed that silly comment;-)

It was this bit that triggered the comment:

    struct {
           ASN.1Cert leaf_certificate;
           ASN.1Cert certificate_chain<0..2^24-1>;
       } X509ChainEntry;

       struct {
           ASN.1Cert tbs_certificate;
           ASN.1Cert precertificate_chain<1..2^24-1>;
       } PrecertChainEntry;


The certificate_chain can be empty but the
precertificate_chain cannot. That's right given the
current spec, but maybe non-obvious.

The 2nd part of the comment was that if you do need
to change the precertificate_chain idea (if the
issuing CA cannot create a precert issuer under itself
e.g. because of a pathLenConstraint) then the
PrecertChainEntry syntax might also have to change.
I dunno if that'd be a real problem now, or only
later, or is just theoretical but I'd say there
will be CAs that can issue TLS server certs but
that cannot issue a sub-ca cert for precertificates.


Cheers,
S.