On Wed, Jul 27, 2016 at 5:52 AM, Ilari Liusvaara <[email protected]>
wrote:

> On Tue, Jul 26, 2016 at 02:29:54PM -0700, Andrew Ayer wrote:
> > On Tue, 26 Jul 2016 23:03:18 +0200
> > Richard Barnes <[email protected]> wrote:
> >
> > > - Add an "identifiers" field to the application object
> > > - Each application MUST have exactly one of "csr" and "identifiers"
> > > - If "csr" is present, then do what's in the draft now
> > > - If "identifiers" is present, then do the same dance, but don't
> > > issue the certificate
> > >
> > > Does that sound sane to folks?  It still seems slightly gross to me,
> > > because of the switching based on the presence of fields.  Anyone have
> > > better ideas?
> >
> > This seems sane, and better than option 1.  The switching is gross, but
> > perhaps it can be made less gross with this logic:
> >
> > - "identifiers" MUST be present.
> > - "csr" MAY be present.
> > - If "csr" is present, its identifiers MUST match "identifiers".
> > - A certificate will only be issued if "csr" is present.
>
> IMO, that makes it even worse, by introducing duplication.
>
> Also, these pre-auth lists aren't actually valid applications,
> and I don't think one should treat those as applications.
>

Yeah, I'm sort of coming around to this opinion as well.  In addition to
this semantic mismatch, it's actually not that much additional machinery to
have the old new-authz endpoint present.  The authorization fulfillment
flow is already defined in Section 6.4, so all you have to specify is that
the client can also use the new-authz endpoint to initiate.

I'll take a cut at a PR.

--Richard




>
>
> -Ilari
>
_______________________________________________
Acme mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/acme

Reply via email to