[ietf-dkim] section 5 rewrite?

[Stephen Farrell]

Eric Allman wrote:
>> Section 5 generally, 1st one is 5.1, 2nd para. This section seems
>> to be all, or almost all, advisory, in which case, it shouldn't use
>> MAY, SHOULD etc. Another example is in 5.3 where it says something
>> that an MUA "SHOULD" do. Suggest lower-casing all the MAY etc.
>> Parts of 5.4 where it says that "From:" MUST be signed etc. is an
>> exception and but would be stated earlier, maybe in around the "h="
>> description.
>>
>> So, is a bit of re-writing of section 5 needed?
> 
> It depends on how extensive the changes are that you are suggesting. The 
> word "re-write" implies that we should through out the current section 
> and start over, to which I disagree.  I'm willing to lower-case some of 
> the terms though and modify individual sentences and paragraphs.

Yes, I meant mostly a lower-casing exercise, and maybe some of the
resulting non-normative text ought be moved elsewhere, to make the
normative stuff easier to get through. (Or maybe not.)

I haven't gone though it sentence by sentence but am happy to sit
down and do that with you in Montreal if it helps.

> I agree with 5.1, para 2, but I disagree with 5.3 --- the point of that 
> is that something somehow SHOULD convert the message to 7-bit MIME 
> before presentation to the DKIM algorithm in order to maximize 
> interoperability.  The converting agent might be an MUA, it might be an 
> MSA, it might even be the MTA, but in any case that conversion SHOULD be 
> done.

I guess I can buy that.

S.