<<< Chronological >>> Author Index    Subject Index <<< Threads >>>

Re: People forging their From: addresses

  • To: Piet Beertema < >
  • From: Ulf Vedenbrant < >
  • Date: Sun, 4 Oct 1998 14:05:06 +0200 (MET DST)


On Sat, 3 Oct 1998, Piet Beertema wrote:

>     And.... If you have enough cpu/bandwidth it's actually almost
>     entertaining to look on the log files when you block SPAM mail.
> Very true... until the entries run in the thousands...

Have seen this happen...;-)

>     The sending host will have to take care of the SPAM mail and
>     it's up to them to enforce a no relay policy.
> Yeah, but how do you enforce that? I've heard too many
> people already who talk about it, report open relays,
> but don't use tools themselves that are at hand, like
> MAPS RBL and ORBS. Sure enough, they're not the perfect
> or ideal tools, but if you don't use them, don't think
> you'll ever get all open relays off the net.

Most SPAM i recieve today is from hotmail,yahoo,att,uunet,aol etc. 
and other rather large email providers. The mail doesn't seems to
be relayed in other means than that the sender probably had an 
account from one of these ISP's. I guess that they are not on
any rbl list. ( haven't checked.. )

Please dont believe that i disagree about figthing SPAM..
I'm just not sure that rbl is the best way to go in all cases..

We got on one rbl-list and wasn't even warned about it. 

It took us a while before we even found out how the relaying
was made due to bad handling of the case from the maintainer
of the rbl-list.

We had a customer that was connected for incoming mail via SMTP and
outgoing via UUCP. ( strange config..;-) 

Of course we should allow relay of a uucp connected customer..
But he also was relaying all incomming SMTP mails to us..
And that was the source of the "open" relay..

I guess that regarding relay problem the X.400 world actually
has something to learn the SMTP world.. The sender normally pays..

/Uffe






  • Post To The List:
<<< Chronological >>> Author    Subject <<< Threads >>>