Download K9 1.10

When Mailwasher became shareware, the program left a big hole in the freeware world. Robin Keir’s K9 fills this gap very successfully. The program does not process e-mail like Mailwasher while it is still on the server, so that bouncing unwanted e-mail is unfortunately not part of the package of possibilities. Instead, K9 acts as a POP server on your own computer or on a network computer (making it possible to use one computer as a spam filter for an entire network). By retrieving e-mail via the server, spam is automatically detected and flagged. After this, a filter in the relevant e-mail program can do with it what the user wants.

One of the main features of the spam fighter is that, unlike Mailwasher which uses a set of rules, it learns what you consider spam and what doesn’t. After making simple corrections a few times, K9 can very precisely separate good e-mail and spam. The program can function after a few days without you noticing. The new version of the program is numbered 1.10. In the new download, which this time claims 99kb of your bandwidth, you can find the following improvements:

• Word parsing now includes ASCII characters 192 – 255 as valid word characters.
• Words are not removed from the opposite database and stats are not updated if the email being reclassified is already of the desired state.
• Added option for only adding words to DB when reclassifying.
• Improved saving of column widths.
• Fixed problem with manual scoring of emails when “…only score new emails” Advanced option was selected.
• Added ability to specify “[Spam]” tag text.
• Spam percentage can now also be added to the custom header when marking as spam.
• Added buttons on the Advanced page for configuring and restoring email program account settings.
• Tray icon changes when Good email has arrived.
• Definable location for “Good” email sound file.
• The program now interprets “encoded” header items (eg =?iso-2022-jp?q?=96=A2=8F=B3?=).

[break]The new version of K9 can be picked up here. The even smaller zip version can be found here.