[SLE] Two spamasssassin issues

From: Hans du Plooy (hansdp_at_newingtoncs.co.za)
Date: 04/10/04

  • Next message: Patrick Shanahan: "Re: [SLE] Two spamasssassin issues" 
    To: suse-linux-e@suse.com
Date: Sat, 10 Apr 2004 14:46:19 +0200

    Hi guys,

    I'm running stock standard spamassassin with SUSE 9.0 Pro. It does not seem
    to be using bayes at all, even though I have specified in both my systemwide
    config and my user config. Summary in the tagged spam says:

    ---- Start SpamAssassin results
    4.80 points, 4 required;
    * 0.1 -- BODY: HTML included in message
    * 0.5 -- BODY: Message is 80% to 90% HTML
    * 0.7 -- URI: Uses a username in a URL
    * 1.5 -- URI: URL contains username and (optional) password
    * 1.9 -- Date: is 12 to 24 hours after Received: date
    * 0.1 -- Message only has text/html MIME parts

    ---- End of SpamAssassin results

    My /etc/mail/spamassassin/local.cf:

    rewrite_subject 0
    use_terse_report 1
    use_bayes 1
    auto_learn 0

    I disabled auto_learn after reading of a lot of people having trouble with
    spamassassin becoming less accurate for some reason that's related to
    auto_learn - I'm still trying to find time to read all the why. I prefer to
    teach my spamassassin by hand too.

    ~/.spamassassin/user_prefs:
    required_hits 4
    whitelist_from *@slashdot.org *@mweb.com *@mweb.co.za *@sterkinekor.co.za
    detailed_phrase_score 1
    defang_mime 0
    dns_available yes
    use_bayes 1
    auto_learn 0

    Both are exactly as they came with spamassassin, I just added my whitelists,
    bayes and auto_learn. This has worked fabulously with all versions of SUSE
    before, I'm not sure what's up now. Am I missing something important?

    Second problem is that a lot more spam slips through lately (absence of bayes
    might contribute) - I see a lot of spam that scores less than 1... Is this
    just a case of the spammers getting smarter?

    Thanks
    Hans 

    -- 
Check the headers for your unsubscription address
For additional commands send e-mail to suse-linux-e-help@suse.com
Also check the archives at http://lists.suse.com
Please read the FAQs: suse-linux-e-faq@suse.com
  • Next message: Patrick Shanahan: "Re: [SLE] Two spamasssassin issues"

    Relevant Pages

    • Re: Amavis-new and Spamassassin Training
      ... >> Spamassassin uses bayes, ... >it rather than relying on the distro to always get it right. ... Viruses & Spam? ...
      (alt.os.linux.suse)
    • Re: For sale Brand New Juicy Couture Sidekick II for $120
      ... the FREE SpamAssassin and the Spam Assassin ... visited the bogofilter site I thought I caught mention of SpamAssassin ... Is there a way to FTP Kmails Spam list, and NonSpam list to the other machines Kmail, so that I don't have to build up a new database from scratch for SpamAssassin? ... and some by Bayes if they include much text. ...
      (Fedora)
    • Re: Spamassasin bribed?
      ... Feed the spam to salearn and watch it evaporate away after awhile. ... I have a long mantra about the RIGHT way to use SpamAssassin ... Use per user Bayes. ... user Bayes has some serious advantages when one user's ham is ...
      (Fedora)
    • Re: For sale Brand New Juicy Couture Sidekick II for $120
      ... Ideally spam should be stopped at source, but I don't suppose there's much chance of that happening. ... BogoFilter authors or the SpamAssassin authors. ... There's no way a BAYES engine alone can do this ... out on rule based flexibility and these BL lists. ...
      (Fedora)
    • Re: Spamassassin and Spambayes
      ... I use kmail - supposedly, kmail detects spamassassin and give you the option of configuring/using it - it provides training buttons on the menu-bar, so you can go through your inbox and separate the spam; I also trained on a number of ham messages that had been filtered into various folders, training on several from each folder. ... Today, I finally got Spambayes running; after initial training on about 200+ messages, I'm already getting around 95% spam detection... ... (And for the most part those train at near Bayes 0.50.) ...
      (Fedora)