From: Greg Long (no email)
Date: Sat Jun 01 2002 - 06:07:54 EDT
I am starting to think little short of violent vigilante justice will
ever stop spammers...all we can do is reduce and try to minimize the
misrouting of legitimate mail.
-----Original Message-----
From:
[mailto:] On Behalf Of Hamish Marson
Sent: Friday, May 31, 2002 2:53 PM
To: Wietse Venema
Cc: ;
Subject: Re: Paranoic Idea to stop spammers
Wietse Venema wrote:
>:
>
>
>>>>I have several ideas that still need to be explored
>>>>(but aren't because of lack of time).
>>>>
>>>>
>>>Aha, you were doing the tests after accepting the mail.
>>>
>>>My test would be done off-line, and no mail from the (client, sender)
>>>tuple would be accepted until there
>>>is a definitive reply. Any reply other than 5XX would
>>>suffice.
>>>
>>>
>>Would you describe the algorithm / tests that you would
>>perform?
>>
>>
>
>It's an idea. I haven't finished the design. All (client+sender) tuples
>would be 4xx blacklisted by default, without using up any storage. New
>verification requests would be queued, sorted and uniqed so that the
>mechanism can't be DOSed too easily.
>
>A (client+sender) tuple would be either 2xx whitelisted for a couple
>weeks after positive confirmation, or 5xx blacklisted for a couple
>weeks after negative confirmation. Only confirmed (client+sender)
>tuples would be be stored.
>
>I would not auto-conform sites when nothing happens. They would remain
>in the non-stored default state of being 4xx blacklisted.
>
>But I may change my mind the next day.
>
>
Nice... I like this...
As long as we can pre-load the white and blacklist...
-- I don't suffer from Insanity... | Linux User #16396 I enjoy every minute of it... | | http://www.travellingkiwi.com/ | - To unsubscribe, send mail to with content (not subject): unsubscribe postfix-users - To unsubscribe, send mail to with content (not subject): unsubscribe postfix-users
|
|
|