Vidiot wrote:
> >far too many newbie list users and administrators set up lists so that replies go
>back to the list. This URL states quite clearly and coherently why this is a bad
>idea. Personally, I have acquired too much scar tissue from misbehaving
> >vacation programs replying to themselves with a reply goes back to the list. It's
>a real nasty death spiral that takes out mail servers on a regular basis.
> >
> >http://www.unicom.com/pw/reply-to-harmful.html
>
> I strongly disagree with this person. I run two mail lists from my server
> and do not have trouble with vacation mail from my list members. I have
> more problems from addresses that go dead, than bounced vacation mail.
> Plus, if I changed the list so that Reply-To was removed, or set to the
> poster, I would get my nuts cut off by the users. The whole reason for the
> lists is to communicate between the members.
then I think we're going to have to agree to disagree. I'm also sorry to hear that
you have clueless users on your mailing lists...
> I quote the following from his document:
>
> "Any reasonable, modern mailer provides this feature. I prefer the Elm mailer.
> It has separate ``r)eply'' and ``g)roup-reply'' commands. If I want to reply
> to the author of a message, I strike the ``r'' key. If I want to send a reply
> to the entire list, I hit ``g'' instead. Piece 'o cake."
>
> I too use elm, exclusively. The "g" key causes a problem, which he fails to
> mention, in that TWO pieces of e-mail is sent out, one to the list and
> one to the author. Doesn't seem unreasonable, except that the author will
> get TWO copies of the e-mail. It is bad enough that I get 300 pieces of
> e-mail a day. I don't need to wade through two copies, especially if they
> don't arrive together. I did a g)roup reply to this, so you will see TWO
> copies of it in your mail box.
sort of. one went to the list digest and the other to me. my filtering puts the
digest into another mailbox that I will read later. The one in my main inbox is
receiving more immediate attention. it doesn't bother me because it matches
my expectations of how mail lists work. I also expect that any news postings may
trigger duel target messages(i.e. one to the newsgroup and one to me).
proper use of filtering tools such as procmail will help make the load manageable.
Trying to work with 300 plus messages a day without filtering shows either
cluelessness or an enjoyment of self-inflicted pain.
> Here is Chip's summary and my responses:
>
> 1) It violates the principle of minimal munging.
> It is still minimal munging. The Reply-To is set to where the message
> came from, in this case the mail list.
this is an arguable point.
> 2) It provides no benefit to the user of a reasonable mailer.
> It certainly does, as it allows me to hit r) and reply back to the
> group.
> 3) It limits a subscriber's freedom to choose how he or she will
> direct a response.
> Bull crap.
au contrair. point 2 and 3 are opposite sides of the same coin. Reply-to prevents
one from easily replying directly to another user. Without reply-to, one can reply
either way to list or to user very easily.
> 4) It actually reduces functionality for the user of a reasonable mailer.
> It has never reduced my functionality.
his point here further emphasizes details of the above points. If reply-to is turned
on, I cannot use simple mailer commands to direct the message to user or to mailing
list. I must instead manually seek out the address from the body of
the message and cut then paste the address. In my world, this means I've lost
functionality. I need to manually do what the mailer formerly did for me.
> 5) It removes important information, which can make it impossible to get back
> to the message sender.
> More bull crap. Since Chris loves the elm mailer, like me, he must
> know about the h) key. That will display all of the header information
> so that he can cut the author's address into a X-buffer. I've yet
> to not be able to find the author's e-mail address in all of the
> headers, from either the lists that I send out, or in the other lists
> that I subscribe to. The From: header is normally still there and
> it contains the author's address. I've yet to be able to send a
> personal reply to an author.
Many of the Windows-based user agents and even some of the Linux ones lose information
about the message sender and make it well-nigh impossible to get back the original
senders information.
> 6) It penalizes the person with a reasonable mailer in order to coddle those
> running brain-dead software.
> I run the same elm software that he does and I don't have a problem
> with Reply-To set to the mail list.
Shakespeare says it best:
"There are more things in heaven and earth, Horatio, Than are dreamt of in your
philosophy."
> 8) It violates the principle of least surprise because it changes the way a
> mailer works.
> Sorry, but I am caught by surprise with the RedHat list because I
> am so used to hitting r) and having the response go to the mail list.
> Now I hit the r), type the message, hit s) to send it and then go
> "Oh Shit!" because it went to the author and not the list.
>
> 9) It violates the principle of least damage, and it encourages a failure
> mode that can be extremely embarrassing -- or worse.
> Bull crap. See #8. After I've said "Oh Shit!", I have to reenter
... Yes, I've seen messages posted to lists that were meant to be sent
> only to the author, but those are few and far between. And yes, I've
> done it.
I think this is true for many people when I say that I have never regretted sending a
message directly to the author alone but have always regretted an accidental list
submission. you can recover from the first mistake but you can never
recover from the second.
assume your users are human. Don't set them up to fail.
> I'm sorry, but Chris' arguments do not hold water with me and cause me more
> problems than it is worth. I guess Chris doesn't mind eating up net bandwidth
> sending two messages out over the net, or eating up disk space on the
> user's computer by sending the poor person two copies of an e-mail that
> should, by all rights, be a single copy.
>
> I vote for the Reply-To being returned to the mail list, pointing to the
> mail list.
like I said at the beginning, we will have to agree to disagree. I have hard-won
experience showing me the wisdom of Chris's arguments, many long-term (decade+) users
agree that reply to munging is considered harmful.
---eric
_______________________________________________
Redhat-list mailing list
[EMAIL PROTECTED]
https://listman.redhat.com/mailman/listinfo/redhat-list
