Tim Bowden wrote,
| A general discussion group is a garden party, and you would not have
| your default response to a speaker in such a setting jimmied so that
| any reply required one to pull the speaker off into a corner of the lot
| so nobody else might overhear. You are entitled to respond, in a free
| and open assembly, in the same manner and with the same volume as the
| original declaration which provoked your reaction.
A person present at such a party is also entitled to think of things to say
that are off-topic or not of interest to the company at large but only to
the one person to whom he or she wants to tell them, and to communicate those
things privately (perhaps after the event rather than during it, but email is
asynchronous anyway). Not all reactions engendered by a speech are relevant
to and suitable for the free and open assembly.
| For other than the discussion forums, mileage varies.
There is a semantic conflict here: "should most replies to a public post be
public or private?" is a different question from "should Reply-To: on a mail-
ing list distribution default to the author or the list?" Chip Rosenthal's
web page addressed only the latter as of the last time I read it.
| Any formula which sends a writer two copies of any and every reply is broke.
"Any and every reply"? Yes, such a formula is broken. Sometimes a direct
copy of a public reply is warranted, but they are sent often when they are
not needed.
But conversely, any formula which sends out to the entire list membership
anything and everything a reader wants to say -- unsubscribe requests, pri-
vate responses that are way off-topic but are triggered because seeing the
author's name reminded the reader that he or she had something to say to the
author that was unrelated to the list, true public responses intended for all
list members to see, or whatever else may come to the respondent's mind -- is
also broken.
So there you have it: both have their flaws. It's a question of which is
worse, and I would tend to think that that must be determined by experience
and history for each list; a blanket rule like "lists for discussion must
clobber Reply-To:" will have exceptions.
In an ideal world lists and MUAs would cooperate on different reply headers
for writing to the list and for writing to the author. Some list software
and some MUAs already support that, I'm told.
