[Spambayes] python.org corpus updated

[Tim Peters]

[Greg Ward]
> ...
> Correct -- I usually put all that stuff in the virus folder, because I'd
> like to see all virus-related junk mail stopped, and I think it should
> be done with different tools from spam detectors.

I don't expect that *this* spam detector is going to do well at viruses
anyway -- although that hasn't been tested.

> Again, my fault for not manually reviewing the spam folder.

Before you do (or maybe while <wink>), let's think about what we're trying
to accomplish:  whether or not you ban various charsets, and whether or not
you look at blacklists, and whether or not etc, all tests run against c.l.py
and python.org traffic have said that a spambayes classifer catches
virtually all the spam and has very low fp rates.

So, at this point, what is the purpose of more testing?  What's the goal
here from your POV?  I can run any number of tests over any number of coming
months, but it's begun to feel simply redundant from my POV.  We're not
learning anything new here, just confirming that this approach works great
for tech mailing lists, and even for python.org's private hobby lists
(provided the classifier is trained on them too).