From: Bruce Lilly (blilly@erols.com)
Date: Tue Jul 06 2004 - 18:27:22 CDT
Charles Lindsey wrote:
> In <40E4C890.2090005@erols.com> Bruce Lilly <blilly@erols.com> writes:
>
>
>>Charles Lindsey wrote:
>>>No, the only situation that could lead to that is when Message-IDs are
>>>omitted in email messages (so you have less things available to refer to).
>
>
>>No Charles, you are wrong yet again. Go read RFC 2822 section 3.6.4 and consider
>>a response to a message with a Message-ID and two msg-ids in In-Reply-To but no
>>References field. Then consider a reply which has multiple "parents" (with
>>Message-ID fields).
>
>
> RFC 2822 makes it clear that it does not address the issue of replies that
> have multiple parents. Our draft also says the same thing. It is
> abundantly clear that the intent of RFC 2822 was that, within the scope it
> attempted to cover, the References-header would ideally contain all its
> ancestors; however, due to the non-availability of message identifiers in
> some articles, it had sometimes to settle for less.
No, in the specific case of a response to a message with Message-ID, no
References field, and two or more ids in an In-Reply-To field, there is no
provision for generating a References field.
> The recommendation^H^H^H^H^H^H^H^H^H^H^H^H^H^Hsuggestion in RFC 2046 was
> quite at variance with that.
RFC 2046 deals with message/partial MIME messages, not non-MIME messages.
RFC 2822 deals with non-MIME messages, not MIME message/partial messages.
They are two orthogonal types of messages.