[exim] Re: spam_score_int signed or unsigned?

2006-04-16 Thread Andreas Metzler
Jakob Hirsch [EMAIL PROTECTED] wrote: Quoting W B Hacker: Steal threads *how* ?? I started a new one. No, you replied to a message in an existing thread, otherwise there would have been no References and In-Reply-To header. Which will cause it to appear as part of the treead in any decent

Re: [exim] sudo - iptables trick

2006-04-16 Thread Tom Kistner
Marc Perkel wrote: Basicly my idea is that when a dictionary tack occurs I want to block the IP address for a short period of time as a load reduction trick with the chain being cleared every few minutes. I've been doing this for a few months with very good results. Not to reduce the load,

Re: [exim] Re: spam_score_int signed or unsigned?

2006-04-16 Thread W B Hacker
Andreas Metzler wrote: Jakob Hirsch [EMAIL PROTECTED] wrote: Quoting W B Hacker: Steal threads *how* ?? I started a new one. No, you replied to a message in an existing thread, otherwise there would have been no References and In-Reply-To header. Which will cause it to appear as part

Re: [exim] sudo - iptables trick

2006-04-16 Thread Tim Jackson
John Hall wrote: On 4/15/06, W B Hacker [EMAIL PROTECTED] wrote: John Hall wrote: [non-interactively sudo'ing from exim to root to add iptables rules] See NOPASSWD in man sudoers. I think I am beginning to see why some folks say Linux is no more secure than Windows. That depends

Re: [exim] linefeeds in check_data acl

2006-04-16 Thread Marc Sherman
Marten Lehmann wrote: I hope my last email didn't sound too angry. I just ment: While use_crlf already exists for routers and transports, can't someone make it available during ACLs, too? I'm sure patches would be welcome. You seem to be the only person who has ever needed this, so it

Re: [exim] Exim domain and login checks for relay

2006-04-16 Thread Jeremy Harris
MrTheo wrote: What I want is that someone who wants to send a mail using the server's smtp services has to login with server's user account, - advertise auth - decide what types of auth you will support. - educate your users in configuring their MUAs - reject, in an ACL, senders who are on

Re: [exim] Exim domain and login checks for relay

2006-04-16 Thread Alan J. Flavell
On Sun, 16 Apr 2006, Jeremy Harris wrote: [...] - reject senders not matching the specific auth data Rejection is not as easy as it sounds. We try to reject mail submissions which present non-existent sender addresses - but mail client software typically does not react well to a 5xx telling

Re: [exim] 4xx or 5xx for notifying overquota condition?

2006-04-16 Thread Mark Nipper
On 14 Apr 2006, Maykel Moya wrote: What error code is advisable to use for notifying overquota condition? I would think a 4xx level error message would be more appropriate for quota issues since it is possibly a temporary situation which may be resolved in the few days a mailer is trying

Re: [exim] Exim domain and login checks for relay

2006-04-16 Thread Jeremy Harris
Alan J. Flavell wrote: On Sun, 16 Apr 2006, Jeremy Harris wrote: [...] - reject senders not matching the specific auth data Rejection is not as easy as it sounds. We try to reject mail submissions which present non-existent sender addresses - but mail client software typically does not

[exim] message-size condition doesn't work

2006-04-16 Thread Marten Lehmann
Hello, at http://www.exim.org/exim-html-4.60/doc/html/spec.html/ch40.html#SECTscanspamass there is the following condition: deny message = This message was classified as SPAM condition = ${if {$message_size}{10K}} spam = nobody I'm using this, but I noticed, that exim passes even

Re: [exim] message-size condition doesn't work

2006-04-16 Thread Stanislaw Halik
On Sun, Apr 16, 2006, Marten Lehmann wrote: deny message = This message was classified as SPAM condition = ${if {$message_size}{10K}} spam = nobody I'm using this, but I noticed, that exim passes even messages with 300KB and 1MB to the spamfilter. Why doesn't it work as expected? My

Re: [exim] message-size condition doesn't work

2006-04-16 Thread Marten Lehmann
Hello, try flipping the '' sign. why? I don't want messages 10K to be spamfiltered. Right now all messages are scanned no matter which size they have. So something is wrong in the condition, but the documentation says, that I don't have to write ${if {$message_size}{10K}{1}{0}}

Re: [exim] message-size condition doesn't work

2006-04-16 Thread Peter Bowyer
On 16/04/06, Marten Lehmann [EMAIL PROTECTED] wrote: Hello, at http://www.exim.org/exim-html-4.60/doc/html/spec.html/ch40.html#SECTscanspamass there is the following condition: deny message = This message was classified as SPAM condition = ${if {$message_size}{10K}} spam =

Re: [exim] sudo - iptables trick

2006-04-16 Thread Walt Reed
On Sun, Apr 16, 2006 at 12:39:49PM +0100, Tim Jackson said: It depends. Obviously if you have mail ALL=(root) NOPASSWD ALL then that's not a good idea, but if you restrict mail to running just some wrapper scripts that invoke iptables appropriately, then it is reasonably secure.

Re: [exim] message-size condition doesn't work

2006-04-16 Thread Stanislaw Halik
On Sun, Apr 16, 2006, Marten Lehmann wrote: try flipping the '' sign. why? I don't want messages 10K to be spamfiltered. sorry, my bad. got it wrong. -- sh pgpv7LIOf6ORE.pgp Description: PGP signature -- ## List details at http://www.exim.org/mailman/listinfo/exim-users ## Exim details at

Re: [exim] sudo - iptables trick

2006-04-16 Thread Marc Perkel
Tom Kistner wrote: Marc Perkel wrote: Basicly my idea is that when a dictionary tack occurs I want to block the IP address for a short period of time as a load reduction trick with the chain being cleared every few minutes. I've been doing this for a few months with very good

Re: [exim] sudo - iptables trick

2006-04-16 Thread W B Hacker
Tom Kistner wrote: Marc Perkel wrote: Basicly my idea is that when a dictionary tack occurs I want to block the IP address for a short period of time as a load reduction trick with the chain being cleared every few minutes. I've been doing this for a few months with very good results. Not