[Spambayes] Problem with POP3 Proxy

[Ian Hlavats]

I am using SpamBayes POP3 Proxy Version 0.4 (February 2004) (binary), with version 2.3.2+ (#49, Oct 23 2003, 15:50:06) [MSC v.1200 32 bit (Intel)] of Python; my operating system is Windows 5.0.2195.2 (Service Pack 4).  I have trained 3 ham and 661 spam.

The problem I am having is that I am unable to remove a previous hammie.db database because I can't find it and this is causing inconsistencies in the Web user interface and possibly in Spambayes' operations.

Hi, the above technical details (re: training) are not entirely accurate and this is part of the problem I'm having. When I first installed Spambayes I tried training it on some spam (661 messages as indicated above) and then I tried to feed it a .dbx store containing more than ham 3000 messages (I had not yet read the docs regarding the 1:1 spam/ham training ratio). After more than one hour of waiting for it to process the ham, I gave up and shut down the server. 

I tried to read how to delete and rebuild the database but I could not find "hammie.db" anywhere on my computer so I could not delete the database. I reinstalled Spambayes numerous times and whenever I opened the Web interface it still showed my previous spam training stats (661 spam, 3 ham). Next I configured Spambayes to use a specific .db file (I gave it the full path this time) and it created the database file and showed my training stats at 0/0.

This was fine, the slate was clean and I was ready to retrain the system. This time I fed Spambayes roughly 600 spam and 600 ham. It has been working great since then, but now whenever I change something else in the configuration settings and go back to the home page, it still shows my old spam/ham training numbers (661 spam, 3 ham)! 

After a short while, if I restart the proxy server it comes up with my current stats (the real ones that I've created since retraining Spambayes), but as you can see from above, it occasionally reverts to the initial training status. Is there any way for me to permamently remove any remnants from my first training database? Since I can't find the database I have no way to delete it.

Aside from this I am quite impressed with your product.

I would appreciate your help,
Ian H.

ihlavats at tarantulaconsulting.com

--
613-265-8251
Tarantula Consulting
www.tarantulaconsulting.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.python.org/pipermail/spambayes/attachments/20040213/e018013f/attachment.html