Possible items for AUTH48

I have now done a careful read through of USEPRO draft-14, and as fas as
the text goes, it seems in pretty good shape (which is not to say that I
agree with everything it says, since it is well known that I have serious
disagreements on two issues).

So here are the niggles that I have found, and that may be worth fixing:

My name, as a joint author, is given as "C. Lindsey". Since I have a
perfectly good middle initial, I would prefer "C. H. Lindsey".

In section 3.5 (Duties of an injecting agent), Step 2, it states:

   "It MUST reject any proto-article ... that has a Path header field
   containing the "POSTED" <diag-keyword>;..."

That is inconsistent with the recently altered section 3.4.1
(Proto-articles) which now states only that

   "the Path header field SHOULD NOT contain a "POSTED" <diag-keyword>;"

Russ pointed out this anomaly a couple of months back, but since both the
Chair and the Area Director flatly refused any further discussion of the
issue, I am not holding my breath for it to be fixed. Neverteless, it
remains an anomaly.

In section 5.4 (The Supersedes Header Field), it is not immediately clear
whether an Approved header is needed, as with Group control messages.
Diligent examination reveals that it is NOT required (and that is
correct). But it might be kinder to add:

   NOTE: It is not required (except in moderated groups) for a Supersedes
   header field to be accompanied by an Approved header field.

In Appendix A (Changes to the Existing Protocols), there should be a
mention of the recently introduce Original-Sender header field.

I have carefully noted all the places where [RFC2822] is mentioned, and it
appears that they would all be equally applicable to RFC5822, which is on
track to become a Draft Standard.



As regards the USEFOR draft, which will also be up for AUTH 48 at the same
time, it would again be possible to update it to RFC5822 under the rules
governing AUTH48. This _would_ involve a change of text by way of a
considerable simplification off the complex (and now unnecessary) syntax
of <msg-id>. There is now only one very minor difference betwee USEFOR and
RFC5822 as regards the syntax of the Message-ID header field, namely that
it is forbidden to contain and '>' (or SP IIRC). That small difference
would be easy to describe - it is intended to permit quick and dirty
parsers of Message-ID to stop reading a message identifier as soon as they
encounter either a '>' or a SP. In practical terms neither of those
currently ever occurs in the middle of a <msg-id> in either News or Email,
so there are no real-world implications of tidying up the syntax of USEFOR
that way.

I would be happy to go through the whole of USEFOR and to docement exactly
that changes would be needed to enable it to reference RFC5822 in place of
RFC 2822. There is no other difference that springs immediately to mind
other than the one mentioned, but obviously it would need to be checked
carefully.

-- 
Charles H. Lindsey ---------At Home, doing my own thing------------------------
Tel: +44 161 436 6131            Web: http://www.cs.man.ac.uk/~chl
Email: [email protected]      Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K.
PGP: 2C15F1A9      Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5