Re: [Shutup] Proposed Charter for something

[Hector Santos <[e-mail]>]

On 12/7/2015 5:50 AM, Stephen Farrell wrote:
>
>
> On 07/12/15 02:34, John Levine wrote:
>> This is increasingly looking like a RG, not a WG.
>
> Hmm, not sure. A good RG is one that'd attract an active
> research community and be relatively long lived whereas
> this seems like a bit of work where we need to do a relatively
> small and discrete amount of research on one topic. (Which
> is roughly "What'd happen if you mucked with received headers
> the MSA passes on to the rest of the mail infrastructure?")
>
> The way I think of it is a good RG would produce a stream
> of academic publications and then some stuff that'd be of
> use to the IETF. This sounds more like one paper's worth
> of research then stuff would be done in the IETF or not,
> depending on the findings.
>
> So I don't think an RG would be a good plan and a WG
> can be chartered to investigate things before proposing
> things.

+1.

I think a working group is sufficient with faster high quality results 
than a longer term research effort.  We have enough 
ietf-industry-man-years, especially in the mail industry, to determine 
what is necessary.

The charter started out with ambitious goals that included 
"well-engineered improvements to the SMTP protocol."

But then it limited itself to specific work items, including DMARC 
related items which is not enough. I believe this effort requires 
liaison with the IETF-SMTP and DNS communities, as well as the DMARC 
and SPF communities.

How about:

   o RFC5321BIS, RFC5322BIS  Working Groups
   o "Email Bible" BCP for Hosting Mail Services Working Group

The latter would produce an Modern Integrated ESMTP guide like RFC1123 
was the "hosting bible" for many internet application hosting 
implementers.

But then again, I can see how the charter may want to keep it simple 
too with the specific work items:

  o New I-D to relax the STD10/RFC5321.Received MUST requirement,
  o New I-D for "Cryptographic Protection of E-mail Headers" (memoryhole)
  o Soliciting new ideas (any I-Ds?) for Loop Detections.


-- 
HLS