On 2006-12-27, Craig Forbes wrote: > On Wed, Dec 27, 2006 at 11:21:17AM -0600, Corey Halpin wrote: > > Also, does anybody know of a way to check if the origin of an email (as > > determined by the earliest of the Received headers) is actually an MX? > > I'd like to avoid sending confirm messages based on email from machines > > that weren't MXes. > > I suspect you will be discarding a good bit of valid email. I think many ISPs > have outgoing mail servers that are not MXs for their domain.
Not discarding. Just putting it into the pending queue. Provided that my Bayesian filter isn't sure about it, and they're not already in my whitelist. Maybe I should be more specific about the problem I'm trying to solve: I want to avoid sending confirm messages out when the From header was "obviously forged". One way to check for such "obvious forgery" (IMO) would be to verify that the sending machine is someone's MX. I'm sure there are other (potentially better) things I could/should be checking. Maybe I should be attempting to valid SPF records? Maybe I should be checking against DNSBL (though that makes me a *little* nervous)? Or should I just give up on my quest to identify "obvious forgery" because doing so is non-obvious? :-) crh
pgpDnr0vav23i.pgp
Description: PGP signature
_____________________________________________ tmda-users mailing list (tmda-users@tmda.net) http://tmda.net/lists/listinfo/tmda-users
