Zitat von Martijn Brinkers <martijn(a)djigzo.com>om>:
This is broken
by design. The message-ID is a MUA thing and therefore
should never be changed by any MTA. If Exchange does not use the
recipient mailbox *and* the message-ID to supress duplicates the
developer have probably not read any RFC. But maybe they have learned
since 2004??
Whether or not Exchange does not right thing in this case is debatable
but that's how they implemented it.
I would rather live with exchange users loose mail (which it does
anyway by different reasons), than to broken e-mail for the rest.
But of course this could be a config setting with a short warning.
From RFC 2822 section 3.4.6
In all cases, it is the meaning that the sender of the message wishes
to convey (i.e., whether this is the same message or a different
message) that determines whether or not the "Message-ID:" field
changes, not any particular syntactic difference that appears (or does
not appear) in the message.
Correct me if I'm wrong but with "sender" they do not mean the person
that created the message. With sender they mean the sender that changes
the messages (in this case the Djigzo server).
As far as i can see the "sender" is the *creator* of the message. With
"creator" being the one who created the content. I would not say that
there is any hint that the actual encoding of the message should alter
the ID. This should apply for all non-MIME --> MIME, 8Bit --> 7Bit or
whatever transformations. The solely purpose of the message-ID is to
give the creator (human) a possibility to include a machine parseable
context-tracker IMMO.
I fail to see any other usage for a message-ID otherwise.
Regards
Andreas