Re: Usefor Status

Simon Lyall <simon <at> darkmere.gen.nz> writes:

> draft-ietf-usefor-posted-mailed-01.txt
> draft-ietf-usefor-message-id-01.txt
> 
> Could the authors (and others) please post to the list their
> opinions on the above drafts. From memory there has been little
> recent comment on them and they (well one of the 2nd and 3rd) may be
> ready for the next stage.

draft-ietf-usefor-posted-mailed-01.txt apparently has a number of
unresolved questions.  Dan Ritter's August IETF 42 notes raise a
number of questions, and I've included my thoughts.

 o Why do we ignore followups-to in mail messages?

Paying attention to Followups-To in mail would then require mail
clients to be able to do NNTP.  It might be very difficult to get
standalone mailers like Z-mail, Eudora, etc., to implement a news
interface, especially for cases where there's a Followup-Host
defined.

Probably better to force all mail clients to work as they do now,
rather than write a spec including a directive that almost no one will
follow.

 o We should add a requirement for handling Newsgroups header in mail 
   if Posted-To header is not present. 

I do not understand this statement.  If the message is posted and
mailed, the Posted-And-Mailed header MUST be set, in which case, the
Newsgroups header MUST be present.  The only time we'd ignore
Newsgroups in mail is when Posted-And-Mailed isn't defined, in which
case, it's an issue of not conforming to the document (probably
predating it), and trying to specify what such software should do is
probably an exercise in futility.

 o Why is Newsgroup header included in mail reply if it's not posted
   to newsgroup?  

I do not understand this, either.  If it's not posted and mailed, it's 
beyond the scope of this document.  (And I don't see anything in there 
in to indicate that replies sent only by mail should have a Newsgroups 
header in it...)

 o Apparently, there is ambiguity in how this has been handled. Some
   MUA/reader agents have interpreted this as "This is a response to a
   news article" and others have thought it to mean "This is also
   being sent via news.  

Right, which is why we have the requirement for Posted-And-Mailed to
be defined before we dare assume what to do with the Newsgroups
header in mail.

I believe that draft-ietf-usefor-posted-mailed-01.txt is complete and
ready to move forward for consideration as a Proposed Standard.

draft-ietf-usefor-message-id-01.txt

I think that this is ready, except that I would like to include an
accurate indication of the probability of collision for the section on
use of only a timestamp and a 64-bit random number.  Before it can be
published, though, we need to also figure out what to do about it in
light of draft-ietf-usefor-msg-id-alt-00.

The two are almost completely redundant, and it seems that it would
make the most sense to merge the two before publishing either.  There
are some nice things in Claus' draft that make sense, but he doesn't
seem to think that use of random numbers is worthy of detailed
consideration.  Alone they're pretty well useless, as are any of the
inputs for the local-part, but combining a 64-bit random number with a
timestamp makes a straightforward way to make a good Message-ID.  I
suspect that this sort of thing is already happening, so I'd prefer to
include some recommendations on how to do it properly.

--

-- 
Matt Curtin cmcurtin <at> interhack.net http://www.interhack.net/people/cmcurtin/