Re: [Exim] use of Resent-* fields

Page principale
Supprimer ce message
Répondre à ce message
Auteur: Philip Hazel
Date:  
À: Jeremy C. Reed
CC: exim-users
Sujet: Re: [Exim] use of Resent-* fields
On Tue, 23 Oct 2001, Jeremy C. Reed wrote:

> What is the expected behaviour of using a Resent-To field?


Things changed with RFC 2822. Exim 3's behaviour is probably bad and
certainly anomalous. I've tried to do better in Exim 4.

> The pine documentation says pine expects to use "sendmail -t" (when using
> sendmail-path: "AND operate in the style of sendmail's "-t" option").


If you are using -t, all bets are off. That isn't an RFC controlled
thing. It's a private interface between the MUA and MTA.

I confess to having completely ignored -t when revamping the use of
resent- headers for Exim 4. I'll check out whether I can find a spec
anywhere.

> So how does Exim know to not send to each recipient and just the
> Resent-To: user?


Exim 3 makes resent- headers supersede the normal headers. This is bad
because there may be several sets of them, and it isn't clear how these
are to be used.

> The exim spec says: "If any Resent- headers exist, it is those headers
> that are checked rather than the original ones." and "If Exim finds any
> Resent- headers in the message, it applies the header transformations
> described below only to the Resent- header set, leaving the unqualified
> set alone."


That's what I thought should be the case, not having thought it through.
In Exim 4, rewrites etc. apply to all headers.

> This is confusing because rfc2822 says "Resent fields are strictly
> informational. They MUST NOT be used in the normal processing of replies
> or other such automatic actions on messages."


Indeed. That's why it has all changed for Exim 4. !!

> Does it mean that if Resent-To: and/or Resent-Cc: is found then the other
> To: and Cc: recipients will not be used with the "-t" option?


Yes, for Exim 3.

In Exim 4, at the moment, To: and Cc: will be used. But if I can find a
specification for -t that says otherwise, I guess I'll have to change
it, even though there may be complications with multiple occurrences of
the resent- fields.


-- 
Philip Hazel            University of Cambridge Computing Service,
ph10@???      Cambridge, England. Phone: +44 1223 334714.