At 1:05 PM -0600 1/8/2000, David W. Tamkin wrote:
> But whenever, in fora for discussing running lists on those
> hosts, someone suggests removing the direct-add feature, other listowners
> howl and scream that their lists need to reach a contingent of people who
> are too helpless with email to carry out the instructions of confirming a
> subscription.
> The objections do not come from spammers but from people running lists for
> non-technical topics, many of whom have actually had experiences with people
> lost and confused at the directions to reply to the confirmation request.
And that's a very true problem.
It's not something I talk about on this list generally, because I
know what the reaction will be, but a number of months ago, I removed
the mail-back validation from a number of my lists. The reason? The
mail-back validation was causing huge numbers of problems, and worse,
significant drop-out of subscribers (on the order of 40%). That was
simply not acceptable.
These days, I get about one spam/slam complaint per 50-75,000
subscriptions. On the mail-back controlled lists, I get complaints
and/or help requests from users at the rate of about 1 for every
300-500 subscriptions. Maybe 1 in 250.
Mailback-validations are an important tool, but they're not the
panacea some make them out to be. They solve A set of e-mail problems
(but potentially create other problems, as the mail-back e-mail
itself can be used as an attack!), but mailbacks cause other
problems. Unfortunately, the people most sensitive to these issues
are also the least open to discussion about e-mail techniques, I've
found. So I long ago gave up trying to discuss it.
We've been very successful in building a system that makes life as
easy as possible for users, both in subscribing and in unsubscribing
-- the "welcome" message, for instance, includes an encoded URL that
hooks into our CGI with the address already in it, so they literally
can get back off with a click. Soon, we'll be encoding every piece of
e-mail that way. And in the same period of time, the number of
'unsubscribe' transactions that requires some kind of manual
intervention (i.e. 'me'), has dropped from 51% of the unsubscribes to
18%.
The keys are (a) make it simple, keep it simple, and (b) reactive
administration. The system requires administration that reacts in a
timely manner as well as systems that work reliably and don't require
significant technical skills.
> Personally, I'd rather see it removed. Anyone incapable of answering email
> is no asset to a mailing list
Sorry, David, but that's a very bogus, elitist attitude. e-mail savvy
does not translate to competency on other topics. It might well be
true on list-managers or majordomo-owners, but why in the heck should
it matter for South Bay Birds or Atlanta-Thrashers?
That attitude just doesn't work well in the reality of a mainstreamed Internet.
FWIW, I firmly believe that mailback validation is a useful tool, and
until someone comes up with a better way of handling it, is a key
technology for discussion lists, especially high-volume ones, because
a slam/spam attack using a few high-volume lists can drown a user
before they realize what hit them. But for smaller, quieter lists,
and for moderated/announce/newsletter type things, in many cases,
I've simply found mail-back validation causes more problems than it
solves.
chuq
--
Chuq Von Rospach - Plaidworks Consulting (mailto:[EMAIL PROTECTED])
Apple Mail List Gnome (mailto:[EMAIL PROTECTED])
Pokemon is a game where children go into the woods and capture furry
little creatures and then bring them home and teach them to pit fight.
