On Sat, 28 May 2005, Bruce Lilly wrote:
Clearly this example shows how 'Reply-To' can be used to suggest
that author prefers replies to go to their smith(_at_)home(_dot_)example
mailbox rather than their mary(_at_)example(_dot_)net mailbox.
Perfectly legal. When there are multiple ways to do the same thing,
and there are no reasons to prefer one over the others...
Agreed.
Where things go wrong is when MUAs try to respond to the following:
---------------------------------------------------------------------
From: Mary Smith <mary(_at_)example(_dot_)net>
To: John Doe <jdoe(_at_)machine(_dot_)example>
Cc: Example list <list(_at_)example(_dot_)net>
Reply-To: "Mary Smith: Personal Account" <smith(_at_)home(_dot_)example>,
Example list <list(_at_)example(_dot_)net>
Subject: Re: Saying Hello
Date: Fri, 21 Nov 1997 10:01:10 -0600
Message-ID: <3456(_at_)example(_dot_)net>
In-Reply-To: <1234(_at_)local(_dot_)machine(_dot_)example>
References: <1234(_at_)local(_dot_)machine(_dot_)example>
This is a mail to you and the list
----------------------------------------------------------------------
Lots of MUAs will reply to both list(_at_)example(_dot_)net and
smith(_at_)home(_dot_)example for 'reply'.
That's what the author suggested for responses. Why do you say
"things go wrong"?
Sorry, but you seem too knowledgeable to not be aware of the
potential problems of using Reply-To in public mails. Do I have to
spell it out? (I thought I already mentioned the Reply-To risk).
Yes, for each individual author who has a mailbox, and who does not
require anonymity. There is of course nothing in the current
specification to prevent that.
If use of Reply-To as a way to indicate ones own mailbox were
deprecated, and instead it were mandated that From should always be
the author's mailbox, then Reply-To would be more suitable for the
purpose I need.
Anonymity is another matter I guess.
Would it be an idea to act on this suggestion?
1. Specifically, more implementations supporting MFT.
Nobody has yet produced a coherent suggestion along those lines that
gained any sort of consensus. Keith Moore identified a number of serious
problems with MFT. So it's highly unlikely.
Could you answer the question rather than my footnote? :)
I have at no time advocated the MFT header, it's presence in my mails
should not be construed as advocacy of it, nor would I advocate it
standardisation. :)
regards,
--
Paul Jakma paul(_at_)clubi(_dot_)ie paul(_at_)jakma(_dot_)org
Key ID: 64A2FF6A
Fortune:
If happiness is in your destiny, you need not be in a hurry.
-- Chinese proverb