[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: SPAM fiiltering - spam breakdown



Alvin Oga <aoga@Maggie.Linux-Consulting.com> writes:
> if  i was gonna tagg an email ( spam ) for later processing ...
> i might as well have spent the 1 second to check it the first
> time and hit the "D" key  ... instead of looking at that email twice

Does your MUA have some concept of folders?  What I do is have all of
my "detected" spam get sorted into a special spam folder; while I
happen to check it routinely, nothing forces me to look at it more
than, say, once a week to look for false positives.  This is probably
just as easy and somewhat safer than automatically deleting "detected"
spam.

>   - my "spam" rules/definition ...
>
>        Deny access from all open relays 
>        use Global RBLs and localized RBLs

To reiterate what previous messages have said about RBLs: you're
basically depending on someone else's arbitrary decision on what's
okay for a mail server -- or a network administrator -- to do.  MIT
has an ongoing problem with RBLs, for example, since it has made a
conscious decision to do something correct for its environment which
disagrees with the RBL maintainers' religion.

>        Reject all emails with "spam content" - not easy to do 

There are a couple of projects that try to do this.  I use ifile
(http://nongnu.org/ifile/), which does content-based filtering, and
I'm pretty happy with it.  (Actually, I use Jeremy Brown's Gnus
integration, but that's nitpicking.)  Doing something content-based
definitely depends on the ability to smack the filter for
misclassifying something, though.  Apparently bogofilter tries to do
something very similar.

-- 
David Maze         dmaze@debian.org      http://people.debian.org/~dmaze/
"Theoretical politics is interesting.  Politicking should be illegal."
	-- Abra Mitchell



Reply to: