[Spambayes] Backup daqtabase

Tim Peters tim.peters at gmail.com
Tue Oct 11 20:02:34 CEST 2005


[Jesse Pelton]
> Yuck.  Don't wanna go there.
>
> I'm pretty sure I've seen discussion on the list of a training regimen
> that involved discarding aged messages (you could even have been
> involved),

Testing frameworks hold on to all messages for as long as a test is
running, and store whatever mappings they need in RAM (like Python
dicts) for the duration; this is a matter of minutes (not months), and
has no UI, no provision for recovering from errors, etc.

> but I guess that would be done by separately maintaining a corpus of ham
> and spam that is periodically used to train from scratch.  Does that make
> sense?

You could do that, and it would be especially easy if using a friendly
mail-store format (like Unixish mbox format, not Outlook .pst).


More information about the SpamBayes mailing list