Re: [Dcrup] draft-ietf-dcrup-dkim-usage review and progress

[Scott Kitterman <[e-mail]>]

If I did things right, -03 should be posted.  Except for the 'should be rsa-
sha256' in the ABNF comment, I took a stab at incorporating all of the below 
items.  I didn't address that one because "sha1" / "sha256" / ... is straight 
out of RFC 6376, so I was reluctant to change it without further discussion in 
the group.

Scott K

On Tuesday, July 18, 2017 04:53:43 PM Murray S. Kucherawy wrote:
> Having talked to a few others here at IETF 99 I'm abandoning my push
> against using MUST NOT with respect to discouraging use of "rsa-sha1".
> Scott, you're free to restore that language at your discretion, and I
> believe consensus on that other thread supports that decision.  Thanks for
> everyone's patience while we worked through that one.
> 
> We're going to appoint Seth Blank as the document shepherd for this one.
> I'll help him through the process.  Unless there's anything controversial
> in here or in the two things on the agenda regarding this document, I
> intend to start Working Group Last Call on it after the session on Thursday.
> 
> I've reviewed the document (as a participant) and I have the following
> additional feedback, in my role as a reviewing participant.
> 
> Abstract:
> - Remove the last sentence.  It's already there in the header of the
> document.
> 
> Section 1:
> - The discussion venue thing should be marked "[RFC EDITOR: Please remove
> before publication.]"  In fact, it's probably a good idea to make that its
> own subsection.
> - This section should also mention that SHA1 is being deprecated (and give
> references explaining why).
> 
> Section 3:
> - "One algorithms" should be "One algorithm" at least; we could also say
> "Two algorithms are defined, but only one is currently supported", and have
> a subsection acknowledging that "rsa-sha1" did exist but it is obsolete and
> no longer supported.  I would prefer the latter.
> 
> Section 3.2:
> - There's an errant comma after "DKIM" near the end of the section.
> - Should we say verifiers SHOULD NOT verify signatures with keys smaller
> than 1024?
> 
> Section 4:
> - The ABNF should allow "rsa-sha256", not "sha256".
> 
> Section 6:
> - Errand double period at the end.
> 
> Appendix A:
> - I would remove the specific draft reference, unless you want this draft
> to wait for that one to be published.  Acknowledging John for a lot of
> source material separately is at your discretion.
> 
> -MSK