ietf-822
[Top] [All Lists]

Re: Message Attributes (considered harmful?)

2004-12-18 03:14:58

At 22.34 -0500 04-12-17, Keith Moore wrote:
annotations seem like they belong outside the message container rather than 
inside.

Attributes added to a message will often not reach all
recipients. For example, "List-"-attributes will only reach
those who get a message through a mailing list which adds
such header, some recipients may get the message other
ways. Resent-headers will only be available to the resender
and those the message was resent-to. Spam-control headers
will only reach those who get the message through the
spam-control filter.

This also means that a person may receive a message more
than once, and with different attributes for the different
versions.

How should this be handled by a receiving MUA? An MUA
handling of such attributes can be:

(1) Do not show them to the final recipient. Most MUAs do
    this for received headers.

(2) There is sometimes a special command for the user to
    tell the MUA to show all such otehrwise suppressed
    headers.

(3) Show them to the final recipient, but not do any
    particular action.

(4) Do some action on the header, such as deleting
    spam-marked messages or putting them in a special folder
    named "Probably spam".

(5) Modify some user action on the message, for example
    "List-" headers may modify the functionality of the
    "Reply-to-all" command.

What should a receiving MUA do if it receives multiple
copies of the same message via different routes, and having
difference in such added attributes? In case (1) above,
there is no problem. In case (2), it may be most useful for
the final recipient to be told of all the attributes. THis
means that the MUA must be able to in some way couple
together different versions of a message received different
ways. Such coupling together is probably also useful in
case (4) and (5) above. For example, if a recipient gets a
message via two mailing lists, then "Reply-to-all" should
allow the recipient to reply to both lists.
-- 
Jacob Palme <jpalme(_at_)dsv(_dot_)su(_dot_)se> (Stockholm University and KTH)
for more info see URL: http://www.dsv.su.se/jpalme/