Re: Skipping RBL checks for internal servers

2015-03-21 Thread David F. Skoll
On Sat, 21 Mar 2015 20:51:49 + RW wrote: > The two calculations produce the same result when > Ns2/Nh2 = (Ns2-Ns1)/(Nh2-Nh1) > i.e. if spam and ham is being added in the same ratio that it occurs > in the database. Yup, that's correct; I got it wrong by extrapolating from a numerical examp

Re: Skipping RBL checks for internal servers

2015-03-21 Thread RW
On Sat, 21 Mar 2015 13:13:13 -0400 David F. Skoll wrote: > On Sat, 21 Mar 2015 15:10:19 + > RW wrote: > > > The only token probabilities that can be skewed by token expiry are > > those than get expired and are then subsequently relearned. > > Yup. But they might turn out to be important.

Re: check a x-spam field

2015-03-21 Thread nobswolf
Am 21.03.2015 um 16:14 schrieb RW: > You need to rewrite the header name before SA sees it so it doesn't > start with X-Spam-. You may want to configure Bayes to ignore the new > header name. ok, this works for me: http://stackoverflow.com/questions/29182749/define-header-rule-for-x-spam-flag-i

Re: check a x-spam field

2015-03-21 Thread Benny Pedersen
On March 21, 2015 12:15:46 PM nobswolf wrote: X-Spam-Status: Yes, score=13.3 required=5.0 tests=AXB_X_AOL_SEZ_S,BAYES_99, BAYES_999,BNM,BNM_IP,DKIM_SIGNED,DKIM_VALID,DUMMY,FREEMAIL_FROM, FREEMAIL_REPLYTO,HTML_IMAGE_ONLY_12,HTML_MESSAGE,HTML_SHORT_LINK_IMG_1, OBFU_TEXT_ATTACH,RC

Re: Skipping RBL checks for internal servers

2015-03-21 Thread David F. Skoll
On Sat, 21 Mar 2015 15:10:19 + RW wrote: > The only token probabilities that can be skewed by token expiry are > those than get expired and are then subsequently relearned. Yup. But they might turn out to be important. > Even then when those tokens are relearned the probabilities will end

Re: check a x-spam field

2015-03-21 Thread RW
On Sat, 21 Mar 2015 11:30:34 +0100 nobswolf wrote: > Hello, > > I use SpamAssassin for quite some time and now I'd like to fine-tune a > little. > > Lately I got a message that was already examined by another instance > of a spam-filter that is not under my control. This message was not > auto-l

Re: Skipping RBL checks for internal servers

2015-03-21 Thread RW
On Fri, 20 Mar 2015 22:08:23 -0400 David F. Skoll wrote: > Bayes expiry is a tricky thing. To do expiry in a way that can be > justified mathematically, you really should expire messages, not > individual tokens. Otherwise, you're skewing the probabilities. The only token probabilities that can

Re: check a x-spam field

2015-03-21 Thread nobswolf
To show my point that this is related to all X-Spam header-fields that are already in the original mail I added a new rule and a similar header to my test-mail: Email-Header: X-SPAM-FLAG: YES X-SPOM-FLAG: YES Rules: header ALLREADY X-SPAM-FLAG:raw =~ /YES/ describe ALLREADY wurde schon m

Re: check a x-spam field

2015-03-21 Thread nobswolf
Am 21.03.2015 um 12:25 schrieb Axb: > > If you can live without the "attached" original msg add to local.cf > > report_safe 0 The report is added AFTER the checks are done, right? I have a rule that is not working. How can change the behavior of the reporting help me here? And: I definitely

Re: check a x-spam field

2015-03-21 Thread nobswolf
Am 21.03.2015 um 12:18 schrieb Reindl Harald: > > > Am 21.03.2015 um 12:13 schrieb nobswolf: >>> can you post such a header section? >> >> After SA there is the original one left in the message which is >> an attachment now. And there is the new one with the >> check-results: >> >> X-Spam-Check

Re: check a x-spam field

2015-03-21 Thread Axb
On 03/21/2015 12:13 PM, nobswolf wrote: Am 21.03.2015 um 11:47 schrieb Axb: This didn't work? : header ALLREADY_MARKED X-SPAM-FLAG =~ /^YES\b/ describe ALLREADY_MARKED wurde schon mal markiert score ALLREADY_MARKED 3 Nope. Just tried. I also tried --lint which gave no results. an

Re: check a x-spam field

2015-03-21 Thread Reindl Harald
Am 21.03.2015 um 12:13 schrieb nobswolf: can you post such a header section? After SA there is the original one left in the message which is an attachment now. And there is the new one with the check-results: X-Spam-Checker-Version: SpamAssassin 3.3.2 (2011-06-06) on nobswolf.info X-Spam-Fla

Re: check a x-spam field

2015-03-21 Thread nobswolf
Am 21.03.2015 um 11:47 schrieb Axb: > This didn't work? : > > header ALLREADY_MARKED X-SPAM-FLAG =~ /^YES\b/ > describe ALLREADY_MARKED wurde schon mal markiert > score ALLREADY_MARKED 3 Nope. Just tried. I also tried --lint which gave no results. and I checked with vi :set list for

Re: check a x-spam field

2015-03-21 Thread Axb
On 03/21/2015 11:30 AM, nobswolf wrote: Hello, I use SpamAssassin for quite some time and now I'd like to fine-tune a little. Lately I got a message that was already examined by another instance of a spam-filter that is not under my control. This message was not auto-learned because the score w

check a x-spam field

2015-03-21 Thread nobswolf
Hello, I use SpamAssassin for quite some time and now I'd like to fine-tune a little. Lately I got a message that was already examined by another instance of a spam-filter that is not under my control. This message was not auto-learned because the score was too low for that. So I got the idea I