SORBS Contact

Steve Atkins steve at blighty.com
Thu Aug 10 04:03:49 UTC 2006



On Aug 9, 2006, at 8:29 PM, Robert J. Hantson wrote:

>
> So with all this talk of Blacklists...  does anyone have any  
> suggestions
> that would be helpful to curb the onslaught of email, without being an
> adminidictator?
>
> Right now, the ONLY list we are using is that which is provided  
> through
> spamcop. They seem to have a list that is dynamic and only blacklists
> during periods of high reports, then takes them off the list after a
> short time...
>
> Or am I just a little naive?

Fairly naive. Spamcop blacklists a lot of IP addresses that send
a lot of email that isn't spam. And some that send zero spam, by
any sane definition.

That doesn't mean to say it doesn't work for you, but don't mistake
a list that'll block a mailserver for a week on the basis of one or
two unsubstantiated reports as _safe_ solely because it will only
block it for a week.

Depending on your demographics SpamCop may have an acceptable
false positive level, but it's not a list I advise most users to use  
as it
regularly lists sources of large amounts of non-spam (such as, for
example, mailservers used solely for closed-loop opt-in email).
Despite that, though, it's quite effective if you're prepared to accept
the false positive rate.

You may want to look at the CBL or XBL if you're interested in a
very effective IP based blacklist with a very low level of false
positives. Not zero, but really pretty low.

Pretty much all the others have levels of false positives that are
bad enough that I wouldn't use them myself, though depending
on the demographics of your recipients they may be acceptable
to you. Using them to block mail to all recipients is likely to be
problematic in most cases. Some recipients who choose to use
it? Sure. As part of a scoring system? Perhaps. Blocking across
all users? Probably a bad idea in most cases.

Cheers,
   Steve





More information about the NANOG mailing list