On Mon, Feb 16, 1998 at 12:45:14PM -0800, Professional Software Engineering
wrote:
And the Message-ID on this message from the original AOL user, it contains
what as the domain portion? I'd lay odds on aol.com unless you have some
goofy sendmail config at gac.edu. Or am I still missing something here?
If the user connects directly to the something(_at_)gac(_dot_)edu SMTP to
post their
message, I would expect it to be different.
What about those users that are dialed-in through a local ISP,
and use that local ISP for sending all their mail, *but* want their
mail to come back to them through AOL?
The problem is that you're proposing solutions without being in
the hotseat. Put your neck (literally, your entire career) on the
line for any one of your proposed changes, and I guarantee that
you'll become a lot more conservative in your proposals.
Not having any examples of the full range of these addresses seen, I can't
say that 4daddy123 might not be one of the spam addresses - in which case
two-numeric limits at either end would fail to catch it. I rather doubt
I'd be seening any messages from people with such goofy addresses.
alpha-numeric-alpha, OTOH, has more opportunity for valid addresses (and
just as many, if not more, for spam too).
The moment you say this, that's when the junkmailers start
changing. You see, they're adaptive. When all numeric addresses
start breaking, they change to numeric plus alpha. When numeric plus
alpha starts breaking, they change to alpha plus numeric plus alpha
(or whatever). Ad infinitum.
Push this to it's logical conclusion, and the only addresses
they'll use to send out junkmail are *valid* addresses, like yours or
mine. Just try handling a few million bounces because your email
address got spoofed for a junkmail session.
--
Brad Knowles _ _
brad(_at_)colltech(_dot_)com |_| C o l l e c t i v e
|_|
http://www.colltech.com |_ technologies _|
"Managing Systems and Networks" [] A Pencom Company []