[Spambayes] using SpamBayes with Cloudmark Spamnet

[Tony Meyer]

> Thanks, Tony.  It does look like the spam score is actually 
> changing.

This has taken me a while, but I finally managed to find time to install
Cloudmark's Spamnet and try this myself.

> Here's what the spam clues said before I clicked "Block":
>
> Combined Score: 69% (0.685541)
> Internal ham score (*H*): 0.0402351
> Internal spam score (*S*): 0.411316
> 
> # ham trained on: 2102
> # spam trained on: 568
[...]
> And here's what it said after:
> 
> Combined Score: 0% (9.49102e-005)
> Internal ham score (*H*): 0.999812
> Internal spam score (*S*): 1.92535e-006
> 
> # ham trained on: 2103
> # spam trained on: 567

There was definitely training between these two processes - one more ham and
one fewer spam have been trained.  Are you certain that this didn't happen
for some other reason?

Using the SpamBayes plug-in from source (that shouldn't change anything) and
SpamNet 2.4.3 with Outlook 2002 SP2 and Windows XP SP1, I can't duplicate
this behaviour.  Neither the "block" nor "unblock" buttons change the "Spam
Clues" report at all.

I think it's reasonably safe to say that whatever is happening, it's not
affecting SpamBayes in any bad way, and so you can just keep on going
ignoring this.  Maybe the version of SpamNet I downloaded is newer, and this
fixes an old issue or something.  Anyway, I'm out of ideas about how to
figure out what this is, so will leave it for now, unless I heard more about
it in the future.  (In which case it's good to have these messages in the
archive for future research).

=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.