[SpamCop-List] Re: Trouble with Help Page

[John E. Malmberg]

In article <casgn3$jel$1 at news.spamcop.net>,
  Michael Lefevre <michael.spamcop at michaellefevre.com> writes:
>
> The page is aimed at people running email systems at ISPs. If someone
> doesn't understand the abbreviations and jargon, the chances are that they
> won't be able to implement the recommendations anyway.
>
> It's talking about spammers hijacking people's computers and then using
> their ISP's mail servers to send spam.  In many cases the spam is blocked,
> but then the ISP generates a bounce message and sends it to the faked
> address.  This is obviously not a good thing.

I like the part about having the user's register what e-mail addresses that
they will be using.  I do think though that the ISP should not rewrite that
information, but add an X- header that tells them were to send the bounce if
the message is rejected.  Only the ISP should be able to decode that X- header
if needed.

The reason for that is that I do (and I am sure others) would not want to make
the other e-mail address public, which is what changing the envelope
information would do.  Much of my e-mail is to mailing lists, and some of them
leave the complete headers intact in places that virms and spambots are known
to harvest.  (And the biggest problem I have with virms is broken virus
scanners)

Is there a page about ISPs and network managers that use generate bounce
messages to faked address instead of ever using SMTP rejects?

And cover about how the external mail servers should verify that an internal
mail server will accept the message before relaying it.

The only time I have seen a double bounce is when an external server accepted
the e-mail, and then an internal server issued an SMTP reject, which the
external server generated the bounce.

While I am sure that the case describe in the FOM does happen, my experience
with gettng the bounces is from mail servers that should have used an SMTP
reject in the first place is more common, and then secondly where external mail
servers do not check for delivery first.

The title of this FOM indicates these things should be covered in it.

-John
wb8tyw at qsl.network
Personal Opinion Only