OT: Spamcop BL - good or dangerous?

John Rudd jrudd at ucsc.edu
Wed Nov 29 12:31:28 GMT 2006


Paul Kelly :: Blacknight Solutions wrote:
> Gerard Seibert wrote:
>> On Wednesday November 29, 2006 at 06:21:54 (AM) Arthur Sherman wrote:
>>
>>> Sometimes I get a message from any of lists I'm subscribed to, that mail to
>>> my address bounces.
>>> And as a reason I see Spamcop blocking sender's (legitimate) server.
>>>
>>> Here comes the question:
>>> What would you use instead of Spamcop?
>>> It gotta be free service, and the more lists the better: right now, Spamcop
>>> is #1 blocking BL in the logs.
>>> I am afraid if I drop it, the blocking will be worse.
>>
>> SpamCop does not block legitimate servers. I use SpamCop myself.
> 
> I'm sorry, but that is complete rubbish. SpamCop users blatantly report
> every and any e-mail they receive even double opt-in mailing lists etc.
> It is an extremely dangerous BL to use if you wish to get legitimate e-mail.
> 
> The only rbl of use (at smtp transaction time) is xbl. Anything else
> will drop legitimate mail, that is a fact.
> 

I'm with you up until this point.

Spamcop is absolute trash when it comes to just about every aspect of 
their operations ... so I wouldn't trust their RBL at all.  At most, I 
might use it in SpamAssassin with a _VERY_ low score.  Even then, I 
would be suspicious of their reliability.

However, I don't think XBL is the only valid RBL to use at SMTP time. 
I've found SBL to be useful, and spamhaus in general to be reliable and 
accurate (not just their XBL).  I therefore expect that I'll also be 
using the PBL (and thus zen.spamhaus.org) in the near future.

The other RBLs are all rather questionable to me for use in blocking (or 
quarantining in MS).  At most, I'd use them for SpamAssassin scoring. 
For that, I also like RFCI (even though lots of other people don't) with 
a low to moderate score.  I'm told MAPS is reliable, but they're also 
expensive, so I haven't really looked at them.



More information about the MailScanner mailing list