When Rich Kulawiec wrote:
> Of course, they could avoid much of the issue simply by requiring
> 100% opt-in with confirmation.
Roger Klorese (so much for my usual practice of using an initial to dis-
tinguish one quotee from another when I include the text of more than one
other person) responded,
| Does that mean that it should be impossible to transfer a list from other
| servers, for instance?
I am not inside Rich's head, but in my own view, it would not be a tragedy
if every member had to confirm wanting to stay on the list when it moved
to a a big public listhost (even if the old server was another big public
listhost). My reasoning is that in such cases the big public listhost's
procedures and policies affect the list and its operations to such a degree
that this is a significant change, and a subscriber's confirmation for being
on the list as it was does not imply consent to being on what the list is
going to become.
If the same ultimate authority will be over the same list with the same poli-
cies and procedures on a different host, that's another story. Then I think
that simply moving the subscription roster is justified.
I'm soon closing my last list (and getting out of the listmaternity game for
good if I have any sense); for the past two years, it has run on the private
site of a member. Twice during this time she has changed upstream providers,
necessitating a change in OS and in IP address but not in domain name, not in
my policies, and not in my software. So I just moved the list; except for a
couple days of downtime it was all transparent to the members. Even before
that, when it moved from its original host to hers, the only thing that
changed was the domain name, and yes, I moved the subscriber roster without
requiring confirmations: the key thing was that it was still a list of the
same topic run the same way. But if she had not come through when I had to
get the list off its original host and I had had to move it to a public list-
host, then yes, it would have been correct to require every member to recon-
firm and bade farewell to those who declined to respond as well as to those
who proactively refused, and if I had just moved the rolls instead, then at
this point I would regret having done so even if there had been no com-
plaints. (I hope there are some people on list-managers who can understand
the idea of getting away with a wrong action and nonetheless coming to regret
having done it despite its having had no ill consequences.)
On the other hand, my list had a sublist (which has already closed; its pur-
pose has expired), and at one point I made such a drastic change to the sub-
list's policies that I gave its members thirty days' notice, telling them
that they had until the effective date of the new policies to send me their
consent to the changes; those who declined to reply and those who objected
would remain on the main list but would be removed from the sublist. None
objected, about 2/3 consented, but the rest did not reply and were removed
(they remained on the main list).
| Or that it should be impossible for an administrator to add a name --
| kicking off a confirmation cycle -- but that names can only be initially
| supplied by mail or from the web, presumably be the requestor but possibly
| by a forger?
As long as that confirmation cycle sends only a terse confirmation request
to the address, and that nothing more is sent there unless a positive reply
comes back, then the unwanted email load of that one confirmation request is
far less than that of an open-ended subscription. Besides a confirmation
request or an unconfirmed (and possibly unwanted) subscription, the only
other alternative I can think of is to ignore all subscription requests and
just never add anybody.
| If I'm allowed to add you but that addition kicks off a confirmation
| cycle, what's the difference if that confirmation message is spelled
| "reply to remove yourself" or "reply to add yourself"?
The difference is all the mail from the list that you'll get if you don't
answer, or if your reply cannot be processed by the bot, or if the mail
transports lose your reply. The difference is that between opt-out and
opt-in, because "take action to remove yourself" and "take action to add
yourself" are equivalent to those, respectively.
| Would it be any different if it said "reply to add yourself" after 100
| lines of list-oriented content?
Yes. That would be worse than saying "reply to add yourself" at the top of a
message whose remaining content was just a terse list description, but not so
bad as one saying "reply to remove yourself."
