At 03:24 PM 11/30/00 +0100, Phillip Helbig wrote:
> > 1. I have a mild preference for "reply" going to the original sender and
> > "replyall" going to the entire list rather than both going to the entire
> > list, but I can cut and paste addresses to make replies go where I want
> > under either system. (The only circumstances when I need to use "forward"
> > is when I really want to forward messages to someone new.)
>
>Note that the addresses of all the listees are not in the message, so
>what you are describing is a bit different than what "reply all" means
>to some people, which means send the response to the sender and other
>recipients. The "all" option cannot know that the list address is not
>an individual address but the address of a list. I think most people
>will not have this option. If the name of the list is in the Reply-to:
>header (or in the From: header), a reply will go to the list, regardless
>of one's email software. I suspect that most replies will not go to the
>original sender, but if so then this name will be somewhere in some
>header (but it will depend on your email software whether it can be
>extracted automatically).
My point was that I do sometimes choose to reply to just the original
sender, so I would be annoyed by the proposed configuration that takes away
the convenient and natural way to do this. I find it unnatural that
"reply" should go primarily to the recipients of the message and not
necessarily to its sender. (I.e., it goes to the send only if he is also a
member of the recipient list. For this list, that should normally be the
case, but need not always be true.)
If I am not misconstruing your points, you complaint about the current
configuration is that "replyall" effectively sends the original sender two
copies of the reply, one as an individual and one as a member of the
list. I understand that some find this annoying, but I happen to find it
less annoying than not having a way to send to just the original
sender. (Since the filters in my mail client are configured to treat mail
sent to me personally differently from the way it treats mail sent to a
list, I actually see that second copy as a potential benefit, but I realize
that not everyone agrees on that point.)
Incidentally, it appears that JICSMAIL is currently omitting the sender of
a message when distributing to a list. If it can do that, I see reason why
it couldn't be made to omit explicit recipients, so the original sender
would receive only the copy sent to him directly.
> > 2. In another thread, someone mentioned all those misconfigured vacation
> > programs that reply to messages from the list. If we set the "Reply-To:"
> > header to point to the list, won't that cause all those bogus autoreplies
> > to go to the list (or require those messages to be filtered out in some
> way)?
>
>Any sensible vacation program should be able to determine from various
>headers if this is a mailing list and not send the message in this case.
>However, looking at the headers of this message, it doesn't seem to be
>possible.
My point was that the other thread indicated that apparently a number of
the recipients of this list either aren't using sensible vacation programs
or are using ones that have not been sensibly configured. Do we really
want to make reasonable operation of this list depend on whethers its
recipients sensibly configure their autoreply generators?
On another front, I received neither an acknowledgement of the message to
which Phillip replied nor a copy distributed through the list, so it
appears we may still have a problem with that aspect of the JICSMAIL
configuration. (I also haven't seen my reply to Van Snyder, but I have no
evidence that any of you have, either.)
--
Kurt W Hirchert [log in to unmask]
UIUC Department of Atmospheric Sciences +1-217-265-0327
|