[spambayes-dev] Interesting unsure

[Skip Montanaro]

    Tim> For body (but not header) tokenization, the option
    Tim> replace_nonascii_chars (off by default) is very effective against
    Tim> junk like this, at least for those whose ham is mostly 7-bit ASCII.
    Tim> That option replaces each "funny character" with a question mark.
    Tim> So, e.g., any oddball spelling for "o" in "love" turns the token
    Tim> into "l?ve";

I'd like to simply strip the accents.  With the current scheme you still
wind up with four related tokens, "love", "l?ve", "lov?" and "l?v?", all
prefaced by "subject:".  Since what the spammer wants you to read in all
instances is "love", I think that's the target we should aim at where
possible.

    Tim> Indeed, my Unsures this week are utterly dominated by trash
    Tim> bouncing back to various webmaster and admin addresses due to the
    Tim> Sobig worm forging sender addresses, like

    ...

Mine too.

    Tim> It occurs to me that I haven't had "a spam problem" since last year
    Tim> -- now I've got "a virus bounce problem" <0.5 wink>!

I just classify them as spam.  It's actually unclear to me why these
"anti-virus" programs feel the need to reply to such messages.  Most of the
time the sender is forged anyway, so the reply goes to someone who doesn't
have the virus.

Skip