Re: [exim] X-originating-IP

Page principale
Supprimer ce message
Répondre à ce message
Auteur: Stephen Gran
Date:  
À: exim-users
Sujet: Re: [exim] X-originating-IP
On Sat, Aug 18, 2007 at 11:03:58PM +0200, Peter Lindberg said:
> I'm getting more and more spam from webmail sites like hotmail, and i
> was wondering if it was possible to write an acl to check the
> X-originating-IP header against blacklists. i thought it would be a
> simple task, but i've failed. perhaps somebody could clue me in :-)


On Sun, Aug 19, 2007 at 07:25:32AM -0700, Marc Perkel said:
> Try this
>
> drop    dnslists = zen.spamhaus.org/$h_X-originating-IP:


On Sun, Aug 19, 2007 at 08:57:10PM +0100, Mike Cardwell said:
>
> In your DATA acl *only* ...
>
> deny condition = ${if 
> match{$h_X-Originating-IP:}{\N^\d{1,3}(\.\d{1,3}){3}$\N}}
>       dnslists  = sbl-xbl.spamhaus.org/$h_X-Originating-IP: \
>                 : bl.spamcop.net/$h_X-Originating-IP:
>       message   = $h_X-Originating-IP: is listed on $dnslist_domain \n\
>                   $dnslist_text


I'm just curious, how do you expect people in say, dial up ranges, to use
hotmail and deliver mail? I realize the original question was a "how do
you do it" type question, rather than a "please debate the merits of this
idea" type question, but I really feel I have to bring some of the latter
question up before someone reads the archives and does something so silly.

I understand the frustration, but really, these webmail services and
ISP mail services are _designed_ for address spaces that shouldn't send
mail directly. If you're going to block the legitimate relay, you might
as well just give up and block hotmail altogether. Now, I do get more
spam than legitimate mail from them, so I might tend to lean towards
that solution, but I want you to realize what you're wandering into.
--
--------------------------------------------------------------------------
|  Stephen Gran                  | Weiler's Law:  Nothing is impossible    |
|  steve@???             | for the man who doesn't have to do it   |
|  http://www.lobefin.net/~steve | himself.                                |

--------------------------------------------------------------------------