[Dovecot] [ot] spam (was: Re: Big sites using Dovecot)
/dev/rob0
rob0 at gmx.co.uk
Tue Sep 26 20:29:08 EEST 2006
On Tuesday 26 September 2006 11:57, Matt wrote:
> > In general, I agree - but once you've seen ASSP in action, you'll
> > never look back.
>
> Does ASSP support vhosted mailboxs? Can I allow each user to
> configure their own white/black lists like spam assassin allows?
This is going way off topic ... okay, not going, it's gone. :) But I
contribute to the problem by tossing out this bit:
I don't see the point in per-user spam controls. Spam is spam. If a
site's sending out UBE, that is a spam site, period. Block it.
Spam is not "mail that $USER doesn't want to see". Spam is UBE. If
$USER subscribed to something and confirmed it before the mailings
started, that's not spam. If $USER happens to be interested in the
stock tips or pharmaceuticals or other such spammer spew, it's STILL
spam, and should be treated as such.
Furthermore, users rarely understand how mail works. They think that
sender addresses really mean something. You got spam from some sender
address, so you should blacklist that address? Well duh, it probably
wasn't really that sender. Maybe you just blocked a real person, an
innocent victim of spammers.
Whitelisting by sender address is just as bad for the same reasons.
Suppose you whitelist an Outhouse Distress user who gets a virus, and
the virus goes out to everyone in the address book.
End users occasionally discover the great FUSSP of C/R systems ... and
thus join the legions of spammers. All out of ignorance. Spam
"solutions" implemented by people who don't understand spam and SMTP
always make the problem worse for everyone.
I'm not paternalistic, at least I don't think so, but I'd like to see
movement away from user spam controls and toward *clueful* server-side
spam abatement. Maybe that would indeed be a FUSSP? No telling, because
it will never happen.
--
Offlist mail to this address is discarded unless
"/dev/rob0" or "not-spam" is in Subject: header
More information about the dovecot
mailing list