On Tue, 13 Aug 2002 15:23:04 -0400 Keith Moore <moore@cs.utk.edu> wrote:
> perhaps, but the blacklists cause the opposite problem - they cause a
> great deal of legitimate mail to not be delivered. IMHO the practice
> of bouncing or dropping mail from blacklisted address blocks is about
> as harmful to the reliability of email as the spam itself.
blacklists vary tremendously in intent and method of operation. it is
probably a bad idea to make a blanket statement like this. lists like
relays.visi.com and sbl.spamhaus.org are very well run, with very clearly
stated intent and method of operation, and minimal collateral damage (i
use both, and have yet to see a false positive.)
some blacklists, on the other hand, see collateral damage as a good thing
and deliberately provoke it.
i don't think anybody ought to be using a blacklist unless they fully
understand it and its mode of operation. while i use certain lists, i also
review my reject logs daily to see what's not getting through -- which is
why i feel confident in my statement about near-zero false positive rates
for the sbl and relays.visi.com -- because i don't treat bls as black
boxes and ignore them -- i pay attention to what they are doing. for a high
volume site, it'd be necessary to roll some tools to cook the logs down --
but i would regard this as a necessary task that would need to be done
well.
furthermore, if an ISP uses a blacklist, it certainly would seem that it is
ethically obligated to make its customers aware of this fact, and it's
implications. it doesn't appear that all ISPs do so.
richard
--
Richard Welty rwelty@averillpark.net
Averill Park Networking 518-573-7592
Unix, Linux, IP Network Engineering, Security