[Mailman-Users] Re: [Mailman-Developers] Reverting question

[Marc MERLIN]

[Reply-To set to mailman-users at python.org]

On Tue, Nov 14, 2000 at 06:13:53PM -0800, Chuq Von Rospach wrote:
> At 5:50 PM -0800 11/14/00, Marc MERLIN wrote:
> >2) If you don't want to take that advise, you certainly shouldn't overwrite
> >    a users' reply to with the list's
> 
> I'll disagree. If you feel you must coerce reply-to, coerce it 
> unconditionally. You've already decided the list's needs outweigh the 

A few others mentionned that.
The problem here, is  that by putting a reply-to, you  are already trying to
prevent me from replying directly to  the poster (well, not quite since mutt
lets me  tell it to  ignore all those reply-tos),  but by munging  reply tos
unconditionally, you are  completely preventing me from  replying to someone
who has a non  valid from address and set a good  reply-to, or someone who'd
like me to reply to him at another address.
Doing  a reply-to  rewrite also  prevents me  from Ccing  a second  list and
setting a reply-to there to move the discussion.

That said, I  could see why others would  prefer unconditionally overwriting
the reply-to.
Well, in both cases, you lose, which  is yet another reason why lists should
not set a reply-to :-)

I  think I  agree with  JC in  saying that  short of  removing the  reply to
misfeature altogether, unconditional rewriting should probably be an option

On Wed, Nov 15, 2000 at 12:27:14PM +0800, Gregory Leblanc wrote:
> Fine, that's all well and good, munging reply to is harmful.  Sometimes
> the harm that it causes is less than the harm that it prevents.
 
I've yet  to have  a single person  show me what  harm not  munging reply-to
creates... It's all about users that have been trained to do the wrong thing
and some of them who then believe it's the only way to go.
 
> This is a good policy in most cases, but there ARE some where it doesn't
> make sense.
 
Yep, the  only one I know  is onelist which  tries to force as  many traffic
back to the list in order to  spam you with more messages which each contain
a small add they make money on.

Ok, more seriously, the only downside of group replies is that for MUAs that
don't support list-reply, like mutt does,  the sender gets two copies of the
answer (which  are easy to get  rid off by removing  duplicate messages that
contain the same message id)
 
> I'm sorry, but this is silly.  If you're going to write reply-to headers
> via your MLM, you need to write them ALL the time, otherwise you end up
> with strange and un-predictable behaviour.  Keep the reply-to header

Point  taken. The  other option  is  wrong  too  though,  for the  reason  I
explained. But I guess if you choose to  shoot yourself in the foot, it'd be
nice  to be  able to  chose which  gun you're  going to  use (i.e.  make the
unconditional rewriting an option)
 
> certainly, but if headers are sending mail back to the list, the user's
> reply-to really doesn't make much difference, unless you need to send a
> private reply.  If you're sending a private reply, you can get it out of
> the x-reply-to header.

Great, I  need to patch my  MUA some more to  attempt to deal with  all this
misconfiguration nonsense.

On Tue, Nov 14, 2000 at 11:42:53PM -0500, Tanner Lovelace wrote:
> Marc MERLIN wrote:
> > 
> > The irony of this coming back over and over again...
> > http://www.unicom.com/pw/reply-to-harmful.html
> 
> One good URL deserves another (straight from the
> "reply_goes_to_list Option" documentation):
> 
> http://www.metasystema.org/essays/reply-to-useful.mhtml

Yeah, except that the author is full of shit.
 
Ok, I'll bite.

- "Reply-To gives the respondant an option which would not otherwise exist:
   namely the ability to reply only to the list."
   false, see list-reply function in good MUAs (mutt is one)

- "If you use a reasonable mailer, Reply-To munging does provide new
   functionality, namely the ability to reply only to the list"
   No comment...
   (the author completely ignores list-reply, and says that MUAs should all
   be patched to allow overriding reply-tos so that they can continue to be
   misused)
   
- "Reply-To munging adds additional functionality, it actually increases
  freedom of choice"
  *cough* bullshit *cough*
  For one, it prevents a thread across two lists and the redirection to one
  list or a third list.

- some mailers apparently do not have a separate reply to all function
  Ok, 1) I have yet to see one
      2) in the point above, the author was saying that any MUA that can't
         override a  reply-to is  not a reasonable  mailer and  shouldn't be
         used. The irony...
  
- "For discussion type lists, I would estimate that ninety percent of the
  time, people want to reply to the list"
  Yeah, so? That's why you have several reply functions, where one does not
  require more work than the other
  (the author mentions having to type addresses by hand, I'm not sure why)

- "Some administrators claim that munging Reply-To headers is harmful because
  it surprises people, and can cause damage when things go awry"
  Damn straight.
  If you reply to the list by mistake due to reply-to munging, you look like
  an idiot, or worse.
  If you reply to the sender only by mistake, no harm is done, the mistake
  can be fixed.

- "It's What People Want"
  Yeah, there are misguided people out there, and they are often vocal.
  Users typically know nothing about mail standards or what header sender
  vs envelope sender means. I don't think there opinion is that relevant :-)
  Ok, more seriously,  users do matter but if they're  trained the wrong way
  when they start, some get very  resistant to change, regardless of whether
  it's a good thing or not.


We did get a  bit off topic here, I apologize to all  the people who already
know about all this, but since we  have lots of listmasters here, I'm hoping
a few with  switch, and hopefully more will configure  their lists the right
way in the first place (that's when it's easy, changing is hard)
  
> There are arguments  on both sides for this issue.   I personally think it
> depends a  lot on  what the  people on your  list prefer.   I run  a large
> non-technical mailing list  where we tried making the  "Reply-To" point to
> the list for a while.   Unfortunately, it resulted in someone accidentally
> posting  a very  unflattering  message  about someone  else  on the  list.

Yep, this happens all the time. Many of us have seen this.

> the "Reply-To" not  point to the list.  For other  lists, however, I think
> it could be advantageous.  Because of  that, I think that letting the list
> administrator decide, rather than the software, is the best method.
 
While  I'm very  much against  reply-to munging  (duh!), I'm for  freedom of
choice. I do think  that mailman should offer a reply-to  option, but with a
very  big disclaimer  and maybe  this discussion  (or another  one) so  that
inexperienced listmasters realize what they're doing.

> I also think, though, that  sometimes setting the "Reply-To" and sometimes
> not,  depending  on  whether or  not  it  has  already  been set  has  the
> potential to confuse  a lot of people.  If you're  going to provide *that*
> functionality, you should make sure it is *extremely* well documented.
 
Agreed.
 
Marc
-- 
Microsoft is to operating systems & security ....
                                      .... what McDonalds is to gourmet cooking
  
Home page: http://marc.merlins.org/   |   Finger marc_f at merlins.org for PGP key