On Sat, Jan 04, 2003 at 09:54:15AM -0800, Bart Schaefer wrote:
3.6.4 says:
msg-id = [CFWS] "<" id-left "@" id-right ">" [CFWS]
id-left = dot-atom-text / no-fold-quote / obs-id-left
id-right = dot-atom-text / no-fold-literal / obs-id-right
Please note that "obs-id-right" is defined (in section 4.5.4) as "domain"
which is in turn defined to include "domain-literal" which does include
bracketed IP addresses.
Hm. It seems my misinterpretation was simply with the format of the
RFC. You're saying that the RHS of the above definitions for id-left
and id-right are in fact to be interpreted as "dot-atom-text OR
no-fold-literal OR obs-id-right". I guess I just never picked up on
that before. :-(
... but the RFC only says
that the "id-right" portion of the Message-ID is MUST be a globally
unique identifier for the message
No, you're misreading. The ENTIRE Message-ID MUST be globally unique:
You're right, I was misreading that too, but a correct interpretation
would seem to render the paragraph meaningless. There is *no way* of
insuring global uniquity without ensuring a portion of the msg-id that
belongs to the network, and a separate portion that belongs to the
server. It's exactly the same with IP addresses, MAC addresses, email
addresses, ISBN numbers and even credit card numbers. They all follow
their respective formats, but unique entities consist of a portion that
is assigned to another body which assigns the rest of the entity.
I believe that RFC2822's assertion that "other algorithms will work" is
misleading, since other algorithms will only work if everbody uses the
same one. In this case, the RECOMMENDATION of the RFC has been accepted
as standard practice, so use of any "other algorithm" will *not* in fact
work.
If the id-left is "sufficiently unique" then the id-right is irrelevant.
There is no such thing as "sufficiently unique". Unique is an absolute.
If you can come up with some proof that you can create an algorithm with
which it is impossible to produce Message-IDs that are identical to those
created by other (unknown) algorithms, I'd love to hear it. ;-)
The problem, of course, is guaranteeing that the left side is unique when
you have no information about what other generators are doing.
Bingo.
--
Paul Chvostek
<paul(_at_)it(_dot_)ca>
Operations / Abuse / Whatever
it.canada, hosting and development http://www.it.ca/
_______________________________________________
procmail mailing list
procmail(_at_)lists(_dot_)RWTH-Aachen(_dot_)DE
http://MailMan.RWTH-Aachen.DE/mailman/listinfo/procmail