Join the AMA (Ask Me Anything) with the Firefox leadership team to celebrate Firefox 20th anniversary and discuss Firefox’s future on Mozilla Connect. Mark your calendar on Thursday, November 14, 18:00 - 20:00 UTC!

Fungování této stránky je z důvodu údržby dočasně omezeno. Pokud žádný článek nápovědy nevyřeší váš problém a potřebujete se zeptat na další řešení, napište nám na Twitter @FirefoxSupport nebo Reddit /r/firefox.

Prohledat stránky podpory

Vyhněte se podvodům. Za účelem poskytnutí podpory vás nikdy nežádáme, abyste zavolali nebo poslali SMS na nějaké telefonní číslo nebo abyste sdělili své osobní údaje. Jakékoliv podezřelé chování nám prosím nahlaste pomocí odkazu „Nahlásit zneužití“.

Zjistit více

I like to create a spam filter based on Message Source when BAYES_99 BODY: Bayes spam probability is 99 to 100%

  • 1 odpověď
  • 1 má tento problém
  • 10 zobrazení
  • Poslední odpověď od Matt

more options

My host is using SpamAsassin. In the Message Source Header Information gives BAYES_99 BODY: Bayes spam probability is 99 to 100%

I like to create a spam filter (in filter settings) based on Message Source (see below) Since filtering on message body is already possible, Seems easy to implement, but How ?


X-Spam-Status: No, score=4.3 required=7.0 tests=BAYES_99=1.9,BAYES_999=0.2, FROM_NOT_REPLYTO=0.5,NO_BITLY=1.7,T_RP_MATCHES_RCVD=-0.01 autolearn=no version=3.3.1 X-Spam-Report: * 1.9 BAYES_99 BODY: Bayes spam probability is 99 to 100% * [score: 1.0000] * -0.0 T_RP_MATCHES_RCVD Envelope sender domain matches handover relay * domain * 0.5 FROM_NOT_REPLYTO From: does not match Reply-To: * 0.2 BAYES_999 BODY: Bayes spam probability is 99.9 to 100% * [score: 1.0000] * 1.7 NO_BITLY FULL: contains a bit.ly url X-Spam-Level: **** X-Spam-Checker-Version: SpamAssassin 3.3.1 (2010-03-16) on mx2.dds.nl

My host is using SpamAsassin. In the Message Source Header Information gives BAYES_99 BODY: Bayes spam probability is 99 to 100% I like to create a spam filter (in filter settings) based on Message Source (see below) Since filtering on message body is already possible, Seems easy to implement, but How ? X-Spam-Status: No, score=4.3 required=7.0 tests=BAYES_99=1.9,BAYES_999=0.2, FROM_NOT_REPLYTO=0.5,NO_BITLY=1.7,T_RP_MATCHES_RCVD=-0.01 autolearn=no version=3.3.1 X-Spam-Report: * 1.9 BAYES_99 BODY: Bayes spam probability is 99 to 100% * [score: 1.0000] * -0.0 T_RP_MATCHES_RCVD Envelope sender domain matches handover relay * domain * 0.5 FROM_NOT_REPLYTO From: does not match Reply-To: * 0.2 BAYES_999 BODY: Bayes spam probability is 99.9 to 100% * [score: 1.0000] * 1.7 NO_BITLY FULL: contains a bit.ly url X-Spam-Level: **** X-Spam-Checker-Version: SpamAssassin 3.3.1 (2010-03-16) on mx2.dds.nl

Všechny odpovědi (1)

more options

Right click the account in the folder pane and select settings. In the junk settings for the account set the option to trust junk mail header set by spam assassin.