[Spambayes] Bayes Outlook Plugin & POP Proxy DBs

[Tony Meyer]

> - is there an easy way how I can move my existing Spam 
> databases across to the Linux based Spambayes pop3 proxy so 
> that I can avoid retraining?

Yes.  You *might* be able to simply copy the default_bayes_database.db file
over and have it work.  Certainly if you're using a pickle, this should
work.  If you're using bsddb (probably the case), then this might not work.
In the source release (i.e. the one you'll have on the Linux system) there's
a script called sb_dbexpimp.py.  This can convert to & from the various
database formats.  Probably the most trouble-free solution would be to run
it on the Windows system (you'll need Python installed), converting the
database to a pickle (or text, but a pickle is probably easier).  Copy this
to the Linux system (calling it .hammiedb if you want), and either convert
it back to some sort of dbm, or just use it as a pickle, depending on your
needs.

> - from what I've seen it seems that the pop3 proxy is 
> proxying to fixed pop3 servers only - is there a reason for 
> this? To support multiple users which use different pop 
> servers I think it would be more flexible to have the pop3 
> userid in the customer's mail client to be of the form 
> "user at pop-server:port" and the proxy can thus interface with 
> arbitrary pop servers 

This is how the (rarely used; mostly example) scripts in the pspam directory
work.  You could modify sb_server to work this way if you wanted to.

I can't recall the reason that pop3proxy.py chose to do it this way (I
wasn't a developer at the time, although I imagine I must have read about it
on the list), but there are advantages either way.  Certainly the way that
sb_server does it makes things much simpler to setup in the mail client (and
simpler to change back and forth).  Unless there are a huge number of
servers that you need to proxy, it doesn't seem that it's that difficult to
list them all in the one option that needs to have them.

=Tony Meyer

---
Please always include the list (spambayes at python.org) in your replies
(reply-all), and please don't send me personal mail about SpamBayes. This
way, you get everyone's help, and avoid a lack of replies when I'm busy.