Re: header-munging

["Randall C. Gellens" <[e-mail]>]

On 7 Aug 1996 16:58:02 -0400, [e-mail] wrote:

> Why can't the MTA do the fixup anyway?  sendmail has been doing this
> for years, and I'm not aware of any problems its caused.

As long as the MTA only does fixups on messages that are being
submitted, and never does fixups on messages that are being relayed,
all is well.  When MTAs "fix" relayed messages, the headers become
misleading.

The draft being discussed (draft-gellens-smtp-submit-00.txt) proposes
the SUBMIT extension as a way of being able to tell submission from
relaying.  It also proposes techniques for guessing in the absense of
the extension.

> What is the MTA to do if the client doesn't
> give permission?  Pass on a broken message that violates the RFC?

If the message is being relayed, the MTA should pass the message on, or
bounce it if it is feeling especially pedantic.  If the message is being
submitted, the MTA should consider fixing it.  The SUBMIT extension is
a way to tell submission from relaying.

SMTP was not designed for message submission.  It was designed for
relaying.  It is being used for message submission.  Knowing when a
message is being submitted allows the MTA to verify and fix the
message, and perform other actions.  Perhaps adding a Sender field if
the From does not match the user (assuming authentication is also being
used).


--
|Randall Gellens             |                    [e-mail]|
|[phone]              | fax [phone] can add ,,,,,,,,6350|
|Opinions are personal;  facts are suspect;   I speak only for myself|