Re: Replies to the list

[Kaz Kylheku]

On Wed, 1 Sep 2010 06:58:49 +0200, Jan Kohnert
<address@hidden> wrote:
> Am Mittwoch 01 September 2010, 06:20:06 schrieben Sie:
                                          ^^^^^^^^^^^^^^

This is broken; in a group discussion, surely this should be:
"Kaz Kylheku shriebt:".

Also, why did you put both me and the mailing list in the To header?

Usually, the list goes into Cc.

>> On Wed, 1 Sep 2010 04:19:56 +0200, Jan Kohnert wrote:
>> > OK, I don't like Flamewars, so I'll stop here. But one thing I have to
>> > add: You are using Mutt, according to your headers. Mutt in fact knows
>> > list-reply,
>> > at least according to [1]. So why not using this? That would solve all
>> > your problems. :)
>>
>> List reply is a poor idea, because it assumes that the
>> sender is subscribed to the list. The assumption is only valid for
>> lists that are configured not to allow posting by non-members.
 
[ snip ...]

> Who is talking about list-only reply??? I'm not.

It was you who mentioned the mutt list-reply above, advising Dave
Rogers to use it.

That is in fact a "list-only reply" function (I checked
the latest online mutt documentation).

Your suggestion that David Rogers should use list-reply, and that it
would solve all his problems, is wrongheaded.

Nobody should use this misfeature, nor implement it
it a mail client.

> lilypond-user sets different mail headers, one of it is called "List-Post".

Yes, lilypond-user does set that header. But, ironically,
I don't see this header anywhere in your message
that I'm replying to just now. Could it be that it didn't
pass through the mailing list, but went directly
to me?

Doh!

Since not all list traffic passes through the list
processor, it's useless to rely on headers that it adds
for purposes like sorting into folders and replying.

The most reliable way to identify list traffic is that the list
address, like address@hidden is among the recipients
(either via To: or Cc:).

> And *you* are free to choose:
> - using reply answers to the OP (or the reply-to he set in his MUA, which 
> would be overwritten by a forced reply-to from the list)
> - using reply-all answers to the OP *and* the list
> - using reply-list answers to the list (so if the OP isn't subscribed, he 
>   won't get the answer)

Sure, you are also free to whack yourself on the
head with a stick; that doesn't make it a good idea.

You're not convincing me that there is value in replying to a
message in such a way that, specifically, the author of that message
may or may not excluded from the reply, yet that message is
redirected to a whole mailing list. Such a function certainly
cannot be regarded as the proper way to reply to list discussions
all the time, and it's probably a bad idea to train yourself into
a habit of using it. Indeed, is it even valid to call it a reply
function,
if the so-called reply doesn't necessarily go to the original sender?
This is more like forwarding than replying.

It's difficult to justify having such reply function supported
in the e-mail client in a prominent way (like, say, a third
reply button). See, people who don't dig deeply into e-mail
client behavior might be misled into thinking that it's the right
way to reply to a list).

It's about as useful as a function which replies only to the users
whose name starts with W, or any other subset which arbitrarily
excludes recipients.