
Signature
Roedy Green Canadian Mind Products
The Java Glossary
http://mindprod.com
> On Mon, 16 Jul 2007 10:17:24 +0100, Martin Gregorie
> <martin@see.sig.for.address> wrote, quoted or indirectly quoted
[quoted text clipped - 7 lines]
>
> It is for detecting forged emails, which are nearly always spam.
There's been a lot of talk about SPF on comp.risks over the years. I
haven't followed the discussions very closely not having much interest in it
but there have been a number of people not particularly happy with it. One
such example:
http://catless.ncl.ac.uk/Risks/23.18.html#subj11.1

Signature
Geoff
Martin Gregorie - 18 Jul 2007 11:22 GMT
>> On Mon, 16 Jul 2007 10:17:24 +0100, Martin Gregorie
>> <martin@see.sig.for.address> wrote, quoted or indirectly quoted
[quoted text clipped - 12 lines]
> such example:
> http://catless.ncl.ac.uk/Risks/23.18.html#subj11.1
My interest in SPF is entirely that it can detect spam that uses my
domain as the (forged) sender. Without SPF there's no way that a forged
sender can be detected and I can do without the backscatter.
As its use spreads its my hope that MTAs will start discarding mail sent
to invalid mailboxes with a forged sender address. Its all they can and
should do: bouncing it just annoys some bystander.

Signature
martin@ | Martin Gregorie
gregorie. | Essex, UK
org |
Malcolm Dew-Jones - 18 Jul 2007 19:20 GMT
: > On Mon, 16 Jul 2007 10:17:24 +0100, Martin Gregorie
: > <martin@see.sig.for.address> wrote, quoted or indirectly quoted
[quoted text clipped - 7 lines]
: >
: > It is for detecting forged emails, which are nearly always spam.
To be precise, it detects 1) does a domain restrict the networks a user
can send mail from while still claiming to be sending mail from that
domain? 2) is a user with a mail address in that domain following that
policy?
SPF by itself does not stop or identify any spam, and it doesn't really
detect forged email - partly because many sites do not supply SPF
information, and partly because an email is not necessarily forged simply
because it comes from a network unrelated to the domain of the mail
address.
What it does do is help a domain to enforce its mail policies by
restricting the networks from which its users can send mail.
The disadvantage to the user is that they must use the correct network to
send mail using that address. The advantage to the user is that if their
domain has a good reputation then their mail may not be blocked.
The advantage to the receiver is that they can reliably choose to trust,
OR NOT TRUST, certain domains.
Spammers often use SPF. The key missing ingredient is many anti-spam
discussions is the recognition that all SPF does is allow you to trust
that spammer domains are sending spam - you still have to list the site as
a spam haven - the SPF itself doesn't tell you anything about it being
spam or not.
: There's been a lot of talk about SPF on comp.risks over the years. I
: haven't followed the discussions very closely not having much interest in it
: but there have been a number of people not particularly happy with it. One
: such example:
: http://catless.ncl.ac.uk/Risks/23.18.html#subj11.1
They discuss "forwarding" of email. Arbitrary forwarding is not
ultimately any different that relaying. Open relays used to be a useful
and cooperative way to pass mail around. That was long ago, and open
relays are now a problem. The policy of forwarding mail is now ultimately
flawed for the same non-technical reasons that open relays are flawed.
They discuss other things too, but I shouldn't spend more time on this.