MFT turns this into a function defined by the sender of the message -
That's a content-free statement. Are you next going to argue that Date
displays don't work because Date is ``defined by the sender''?
What I'm saying is that, when both MUAs fully support Mail-Followup-To,
the behavior that's by far _most commonly desired_ in each situation is
also the _default_:
* When someone joins a mailing list and then sends a message to it,
he _usually_ doesn't want a copy of the followups. In the
Mail-Followup-To world, _by default_, he doesn't get a copy.
* When someone sends a message to a mailing list without joining it,
he _usually_ wants a copy of the followups. In the Mail-Followup-To
world, _by default_, he gets a copy.
* When someone wants an extra copy of the followups to his message,
he usually also wants an extra copy of the second-level followups,
third-level followups, etc. In the Mail-Followup-To world, _by
default_, this is what happens.
* When someone sends a message to several mailing lists, he _usually_
wants the followups to go to all of the lists (and him if he isn't
on any of them). In the Mail-Followup-To world, _by default_, the
followups go to all of the lists.
Everything can be overridden by the recipient. Everything can be
overridden by the sender. But this manual work is drastically reduced
when we get the defaults right.
---D. J. Bernstein, Associate Professor, Department of Mathematics,
Statistics, and Computer Science, University of Illinois at Chicago