[Greylist-users] Central whitelist database

[Eirik Oeverby]

Hi folks,

I have posted here earlier about the possibility of a qmail port, where
I mentioned that I am running a few mailservers with quite a lot of
traffic. After following the threads here for a while, I'm still pretty
certain I'm going to give greylisting a go - once I can get it working
with qmail. Any qmail-folks aorund? :)

However there are many things that can be done before it gets to that.
As a preparation, I have been toying with the idea of setting up some
kind of central whitelist, a database that can be queried and/or
subscribed to by people running MTAs and who wishes to make the burden
upon their users as small as possible.

It is an inevitable fact that as people start implementing greylisting,
there will be a lot of duplicate whitelistings among the hundreds or
thousands (or more) mail servers using it. For example I'm pretty
certain that after a short while, practically everyone will have sites
like AOL, T-Online, probably even Hotmail etc. in their whitelists, in
one form or another.

Now .. Why shouldn't these be whitelisted from the get-go? And all the
other smaller sites that are 'confirmed good' by someone else? After
all, the absolute majority of people implementing greylisting will do it
for the number one reason of getting rid of spam - and if some spammer
somehow manages to get his or her domain whitelisted in such a central
db, it will quickly enough be discovered and removed.

I can even think of a mechanism where this central greylisting can
happen automatically. For instance, if mail from a certain domain is
coming in often, and is always being let through over a period of time,
it could be assumed that this domain must be a good one, since noone
complained about it or blacklisted it. An MTA subscribing to the
'central' whitelist could then submit this domain for acceptance into
this central database, where some kind of validation would happen (i.e.
checking that this domain has also been submitted by others) and
eventually it would be put into the database.

I stayed halfway awake most of last night half-thinking, half-dreaming
about this (yea I know I'm sick ;), and while I see certain drawbacks
and perhaps even some ethical issues about it, I think it is doable.

If it is interesting, I can even provide a server where this can be
developed and implemented, atleast for the prototyping and initial
testing. If it ever bears fruit, I'm afraid the traffic might grow
beyond what I can currently handle (I'm on a triple-homed ~2gbit
back-end, but capped to 4mbit at the moment).

Let me know if this is interesting to any of you. I do believe this
might be one of the really major contributions to the war against spam.


/Eirik

-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 187 bytes
Desc: not available
Url : http://lists.puremagic.com/pipermail/greylist-users/attachments/20030624/79447d7c/attachment.bin