On Sat, Jul 22, 2006 at 09:10:12AM -0700, Marc Perkel wrote:
> Martin A. Brooks wrote:
>> Marc Perkel wrote:
>>> When I get to the point where I need security I will implement it.
>> That, for me, is sufficient nails in the coffin for your project.
> I'm looking for people who want to find ways to make it work, not people
> who are looking for ways to make it not work.
Marc,
What Martin is saying, and it's something I agree with, is that many of
the worst bits of security design and attempts to patch things up have
come from not considering the security and trust implications right from
the beginning. SMTP, for some, is a case in point; they say "we have to
totally redesign SMTP from the ground up to make it ``secure''", I'm sure
you've heard that being said. In some ways, the situation we are in (wrt
spammers, joe-jobs etc) is because the designers of SMTP *didn't* think
about the possibility of abuse, although more probably, they weren't
thinking about abuse in the context of deployment at the scale to which
it has been deployed.
If you are saying "I'll bolt on this security-thing later", then you're
fundamentally misunderstanding the security design point of view, and I'm
inclined to agree with Martin in his assessment here.
For any black or white or any other colour list, there is an element of
wanting to know how they are doing their DoS resistance, any time one
has got big, they have been hit by major attacks from the spammers they
are blocking. ORBS gave up, some years ago, after too many of these
attacks, and when they did, decided to list everything in order to get
people to stop using their lists. This was obviously pretty broken
behaviour for anyone trusting them.
One of the questions that for me, you've failed to answer, is:
What happens when someone lists one of your "partner" server operators,
does that change their rating?
And another, that comes to mind because I live in a country with stronger
Data Protection laws:
If I'm feeding information about every message (spam/ham) and source to
you (which I presumably need to do in order to make your statistics
useful), and you obviously know the destination, because it's your peer
host, then suddenly you have some quite powerful mail-flow information,
too, which could come under some of the otherwise dreadful RIPA in this
country at least.
Sorry to not be entirely supportive of your ideas, but constructive
criticism, which is what I hope some of this is, is what will help you
produce a better idea next time. "Those who do not learn from history are
doomed to repeat it" as a corollary to some implementations of "thinking
outside the box".
Cheers
MBM
--
Matthew Byng-Maddick <mbm@???> http://colondot.net/
(Please use this address to reply)
