** Sometime around 2:31 -0500 11/19/98, Chuq Von Rospach sent everyone:
>At 11:29 PM -0500 11/18/98, Vince Sabio wrote:
>
>> Just for the record, I disagree with nearly every point made in that
>> treatise. Overall, it is the list-owner equivalent of the GOOD TIMES
>> virus hoax.
>
>and just for the record -- I disagree with Vince. While the document
>and I don't agree 100%, I find it very persuasive and right-on.
And just for the record, I disagr-- oh, never mind. <g>
>Here's how I look at it. From the point of view of the naive/new user,
>the Reply-To is impossible for them to over-ride. They don't know how,
>and that only encourages chaff and accidental posts, not to mention
>that users who want to reply privately can't.
(Oh no ... it's ... it's ... it's ... .it's happening
<RANT MODE>
... again! *sigh* Sorry about this ...)
Okay, let's take an example list: Mac-L. (I think you're familiar with
that one, Chuq; for anyone else, it's a Macintosh discussion list that
seems to attract people who can't even tell which box is the modem. Some
of the users are really sharp, but we have lots of folks who are very
new to computers, not just the'Net.) Mac-L is configured as REPLY-TO-LIST,
and it is a VERY rare occasion (like once/month or less) that someone
sends an "accidental" post to the list. To be honest, I can't even recall
when the last one occurred. But the extra convenience it provides them
seems to be well appreciated -- whenever the subject comes up, there
are always folks who contrast Mac-L with "those other lists" (REPLY-TO-
SENDER configured lists), which they clearly find more difficult and
more confusing to navigate.
Mac-L is also the list on which I've conducted most of the REPLY-TO polls
(three of them to date). Maybe it's just my users, but they definitely
seem to prefer the REPLY-TO-LIST configuration. (There are those who do
not, but they are far in the minority, poll-wise.)
>On the other hand, users who want to go reply-to-all as default can. I
>do, for instance. By *not* coercing reply-to, I leave the choice in the
>hands of the author of the message. I find it very arrogant to assume I
>know better than the author how a message should be sent, especially
>since by coercing the reply-to, I make it DAMN hard for a user to
>change it back, while if I don't coerce it, it's fairly easy to do the
>opposite.
If the choice is merely the default of whether it goes to the list or
goes to the author, then I don't find it to be at all "arrogant" for me
to decide that replies should go to the list -- especially if that is the
expressed preference of my subscribership. The only exception to this
is when the person composing the reply WANTS to send it to the author,
AND the author has a Reply-To that is different from the address on his
>From line. Obviously, Reply-To munging blows away the original Reply-To
(or relegates it to an X-Reply-To (or similar) line, where the sender
will probably never see it), so the the author must place his preferred
address in his .sig, AND the sender must be smart enough to look in the
.sig and use that address. In my book, this is the only disadvantage of
Reply-To munging, and in the years that I've been running mailing lists,
I've yet to have a single complaint lodged about this. Thus, I tend to
think that it's a very rare occurrence (if it occurs at all).
>> This includes
>> the original author in the reply. When the NEXT person does a "group"
>> reply, the original author AND the secondary author are included. And
>> all THREE of them are included on the third reply. And so it grows.
>>
>> This is a PITA.
>
>it's actually a very tiny one. I've studied this -- the amount of
>"duplicate" mail people get is NEVER even close to 1% of a person's
>total e-mail traffic.
At least one of the non-munged lists that I'm on has a tendency to grow
the To line nearly without bound. I find myself (as others have)
receiving direct mail long after a thread has moved off to a different
topic. It's not unusual for someone to post a "please take me off the
replies to this thread" message to the list. Several of us have begged
the list owner to force a Reply-To <list>, but for whatever reason he
prefers to keep the list configured as it is.
That's merely an exmaple. There are cases where REPLY-TO-SENDER is
overwhelmingly the preferred configuration, but somehow I don't think
I need to defend those cases. ;-)
>And while some individuals are rather sensitve to
>these duplicates, most folks DON'T CARE. I'm sorry if you're one of
>them -- but it's not significant and it's not a real PITA.
See example above. While that list is the best (worst?) example, there
are others. I think that the larger the list, the less probable that it
will be a problem (oddly enough) (well, not oddly at all if you think
about it). So I'd say that it's list (and subscribership) dependent.
>And forcing
>ALL users into this mode because a few people get upset once in a while
>is a bad reason to coerce reply-tos.
Ah, but NOT forcing it tends to upset even *more* users (in most cases
IME). Using the minimum-user-upset metric, we arrive at Reply-To munging
as a recommended configuration in many cases.
>>>When I started running email lists, I munged 'em all. One day I
>>>accidentally sent a private, personal reply out over one of my own damn
>
>> Ah yes, the testimonial to personal stupidity. How touching.
>
>Vince, don't get demeaning.
Too late. :-)
>you only need to send one really, really
>embarassing (or damaging, dangerous, libelous or just plain stupid)
>piece of e-mail to realize that ONE is enough.
BTDT, and I *learned*. I did not force a new list configuration down my
users' unwilling throats because of my own stupidity. Instead, I tried to
do what Darwin would have done in this situation -- I learned a new skill.
>If you've never done it,
>bless you. May you never. but if you do, maybe you'll understand a
>little better and not pull this "how touching" stuff. Phttt. You're
>above that kind of snide insult.
Am not. ;-)
>> Overall, it's a personal preference. I've held several "polls" on my
>> lists (one in particular) in response to suggestions that I change the
>> list to be, in ListProc parlance, REPLY-TO-SENDER instead of REPLY-TO-LIST.
>> In each case, the poll results have been OVERWHELMINGLY in favor of
>> REPLY-TO-LIST.
>
>I've not only done polls, but I've done user testing and studied how
>users USE the lists. They may prefer the coerced reply-to (but
>frankly, I've run polls, and I can guarantee you I'll get whatever
>answer I *want* simply by how I state the question)
So can I (I think), but I really try to keep the polls as objective as
possible. I think it's important to realize that we're there for the
subscribers as a whole; if we manage the lists for ourselves, and that
happens to be substantially different from the desires of the
constituency, then that consitituency will simply go elsewhere.
>but my user
>testing and my mail-flow tests both strongly indicate the reply-to
>causes many more problesm for the typical user than it solves.
My *experience* (and I've run both flavors of lists) (still do, in fact)
indicates that Reply-To is not only preferred by the majority of users,
but works out VERY well overall. Again, maybe it's the lists that I run,
but I'd be surprised -- I *know* that there's a lot of subscriber overlap
between your lists and mine.
>> I realize that I hold a minority position on this (or else the REAL
>> minority is very vocal on this issue), but hey, SOMEONE has to stand up
>> for the virtues of Reply-To munging, dammit.
>
>be my guest. Just understand that you're wrong.... (grin)
In all seriousness, though (okay, I know you were serious, Chuq, but
just go with me on this <g>), I really think that it's a matter of
preference. In many cases, it's the list owner's preference. In some cases,
it's the subscribers' preference. Might be a combination of the two. But
one of the things that I definitely find disturbing in the anti-munging
treatises, and in the anti-munging camp overall, is this tendency to
ignore the preference of the subscribership. If your subscribers really
prefer REPLY-TO-SENDER, by all means, do it. But if you want to talk
about arrogance, then assuming that the list owner knows what's best for
his (or her) constituency -- even if it that runs counter to the
preferences of the subscribers -- is the epitome of list-owner arrogance.
So, in my book, it really comes down to what the users *prefer*. Which is
why I run both REPLY-TO-SENDER and REPLY-TO-LIST configured mailing lists.
Rather than decry one or the other, why not just leave it up to the users?
</rant mode>
Now stop disagreeing with me over here and go subscribe to the *new* OAB
list, Chuq -- so you can disagree with me over THERE. ;-)
- Vince
P.S. -- I appreciate the kind notes of support on this subject that have
been sent to me by personal mail. Too bad we don't have reply-to munging
on this list, since it would have been nice if they'd accidentally been
sent to the entire list. <heh heh heh...>
