Re: Reply in mailing lists

[Brian May <bmay@csse.monash.edu.au>]

>>>>> "Per" == Per Abrahamsen <abraham@dina.kvl.dk> writes:

    Per> Brian May <bmay@csse.monash.edu.au> writes:
    >> I think the To field will prevent the news message from being
    >> posted (unless I am mistaken).

    Per> You are mistaken.

I am...? I will take your word for it ;-)

I assume then that if Gnus sees a "newsgroup" header, it will always
post it?

Similarly, if Gnus sees a "To", "Cc", or "Bcc" header, it
will always post it (independently of if it mails it or not).

    >> IMHO, "mail-copies-to" should override the setting for
    >> "to-address", but would it?

    Per> Try it.

Amazing, it worked! Assuming my test wasn't flawed... (I don't
think it was).

So now I can summarize for mailing lists (and I assume the same holds
for all newsgroups) that have the "to-address" set:

narrow reply
- reply only to author - this is obvious

wide reply
- reply to everyone specified in original message (this may/may not
include the mailing list).
- except if original mail contains "mail-copies-to: never"[1],
in which case the original author will not get a separate copy.
Not sure about other people.

followup
- same thing as "wide reply" for mailing lists where "to-address"
is not set. Otherwise...

- reply to "to-address" only.
- except if "mail-copies-to: always" was set, in which case a CC
is sent to the author.
- if you want to manually force replying to the original author,
then pushing C-c C-t (did I get that correct?) will add him/her
to the "to" field.


So, me thinks it is a pity that the wide reply wont always include
the address of the mailing list, but everything else seems
rather sane and straight forward.


Note:

[1] I can't remember if this is never or nobody. Please substitute for
correct value, and don't take my version as being up-to-date and
correct.

-- 
Brian May <bmay@csse.monash.edu.au>