pkgsrc-Users archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: Off-list reply taken back on-list involving check-file-prefix crud and pkg_add not replacing by default.



"Thomas Mueller" <mueller6725%twc.com@localhost> writes:

> How are NetBSD and FreeBSD lists configured to reply to the OP and cc:
> the list?

They aren't really, but they are :-)   They do not insert a Reply-To:
header, and this is correct behavior according to the mail RFCs.
Therefore, when a message arrives, one can use one of two MUA actions:
  reply: compose a message to the sender
  reply all: compose a message to the sender and all recipients

Part of the point, in the view of those that think the RFCs are correct,
is that "reply" should never send content intended to be private to a
list.

> It's a matter of convention, and depends on whether the sender of a
> message adheres to that convention, which I am not doing.

It's entirely reasonably for someone to reply to list only, if they choose.

> Most other lists' protocol is to reply to list only.

There is a difference between a convention to stay on list for content
that belongs on list, and  the list being configured to insert a
"Reply-To: list_address@" header.  That header causes the MUA reply
action to compose a message to the list.  This is wrong because the
specs say that Reply-To may be used to say where the *sender* wants
replies, and dangerous because private messages get published.

As for protocol as a social construct, I often send hints to people, and
essentially always do that on list.  Sometimes I get private followup
replies, and usually, when I think the content would serve the other
readers, I tell the person that I only provide on-list help.

> When I respond to an off-list reply, I may respond on-list without
> saying anything about off-list reply and certainly not making it
> blatant in the subject line.

That seems ok to me.  The real issue is publishing text that was sent
privately.


Home | Main Index | Thread Index | Old Index