OSEC

Neohapsis is currently accepting applications for employment. For more information, please visit our website www.neohapsis.com or email hr@neohapsis.com
 
From: Jay Maynard (jmaynard_at_conmicro.cx)
Date: Sun Dec 01 2002 - 18:14:20 CST

  • Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

    On Sun, Dec 01, 2002 at 06:04:27PM -0600, Russell Mosemann wrote:
    > This is not such a good idea. The solution would be to put SpamaAssassin
    > in front of postfix, and that would hobble postfix to a large extent.

    That's why I want to insert SA at the same place in postfix processing as
    the body_checks filtering. I don't want to front-end Postfix, for all of the
    reasons you cite. (BTW, I've been more successful with stopping spam with
    the smtp_* filtering than I ever was with sendmail.)

    If I wanted to simply drop spam on the floor, I could do that now with a
    little bit of kludging. I don't. I want to abort the transfer of the spam.
    The body_checks filtering does that, but it's both inflexible and
    considerably less good at catching spam than SA. (I don't particularly want
    to reinvent spam detection; that's why I got SA in the first place.)

    As I think about it, perhaps the best way to do this would be to have the
    message run through SA (in the form of spamc/spamd, to lessen overhead)
    right before the "250 OK: queued as <id>" message is sent; the return code
    would then determine whether the message is queued, or else rejected. This
    would seem like a pretty simple hook to add, if it's not there already.