Errors-to: owner-drums@CS.UTK.EDU
Date: Thu, 27 Jul 1995 10:38:03 +0200 (MET DST)
From: Jacob Palme <jpalme@dsv.su.se>
To: ietf-drums <drums@CS.UTK.EDU>
Subject: Differences between messages with the same Message-ID
Message-Id: <Pine.SUN.3.91.950727102930.10022A@ester>
Mime-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII

Harald.T.Alvestrand@uninett.no writes:

    2.2.  The identified object

    The message-ID locator resolves to an Internet message.

    This identifies a message header and a message body conformant to
    RFC 822 and (possibly) MIME; due to the history of Internet
    messages, the following differences must be expected between
    different copies of the same message:

     (1)   Variable number of "Received:" lines in the header

     (2)   Randomized order of other header lines

     (3)   Addition of "Resent-*" headers

     (4)   Addition/subtraction of other non-standard headers like "X-
           Listproc-version", "Old-Received" or "PP-Warning"

     (5)   Changes of content-transfer-encoding for MIME messages

     (6)   Changes of wrapping for multiline headers

    An object is still considered the same object after undergoing
    these changes.

    The following changes may arise because of errors or misfeatures
    that are known to occur in Internet mailers, but represent
    erroneous representations of the message:


     (1)   Changes to, addition of or removal of standard headers like
           "To:", "Subject:" or "Content-type:"

     (2)   Substituting tabs for spaces or vice versa

     (3)   Adding or deleting blank lines at the end of the message

     (4)   Breaking of lines without using a MIME content-transfer-
           encoding

Comment: It is a very good idea to specify somewhere in an Internet
standard which differences can be expected between messages with the
same Message-ID. However, I suggest this is included in the new
rewritten 821+822, rather than in a document defining MID URI-s,
since the information is valid also outside the URI usage area.

In addition to the reasons Harald lists why two messages with
the same Message-ID can differ, here are some additional such
reasons:

     (7)  The existence and content of Bcc headers can vary between
          copies of a message with the same Message-ID.

     (8)  If the message has a header field of type "Incomplete-Copy"
          (defined in RFC 1327) then one or more body parts can
          be missing.

     (9)  If the message has any of the header-fields "Conversion:"
          or "Conversion-With-Loss:", then the message can differ
          in character set encoding from another version of the
          message with one of these two header fields, or from
          a version of the message without these two header fields.

------------------------------------------------------------------------
Jacob Palme <jpalme@dsv.su.se> (Stockholm University and KTH)
for more info see URL: http://www.dsv.su.se/~jpalme