[BozemanLUG] Best spam filter

[Jordan Schatz]

> But what we need is better identity control.
That would work, but that is alot of work...

I think the problem is artificial intelligence. A human can quickly identify
spam, but from what I've seen most software is STILL bad at it. That is what
needs to be fixed. Clearly some companies (Google, MailRoute) have built the
software, but it looks like the open source stuff is behind... If anyone wants
to fund me I bet I could fix that with a year or two of work : )

- Jordan

On Fri, 16 Nov 2012 10:10:38 -0700, "David Eder" <david at eder.us> wrote:
> I've not responded before, because my input is probably not very useful. 
> But what we need is better identity control.
> 
> Here is what we have now:
> 
> We give company A our e-mail address.  Company A is well behaved.
> We give company B our e-mail address.  Company B is not well behave and
> abuses our trust.
> 
> We cannot tell which company is the bad actor.
> 
> This is what we need:
> 
> We give company A an e-mail address unique to company A.  We give company
> B an e-mail address unique to them.
> 
> When we get spam, we know the bad actor and we simple shut off that access.
> 
> I accomplish this for myself by running my own server and have an alias
> file that is hundreds of lines long.  I have been somewhat inconsistent in
> the past, but generally it works really well.  I can tell you, for
> example, that Walmart is *much* more trustworthy than Target.
> 
> We probably need something like {pick your favorite password manager} for
> e-mail addresses along with a catchall with white/black list.
> 
> If you have multiple customers, you could just tag it on.  For example
> bob at mycompany.net might have bob.*@mycompany.net.
> 
> This of course will fail on mass adoption as it make e-mail buckets
> guessable.
> 
> David.
> 
> 
> >
> > I run a few email servers, and spam has been quite a problem. Any opinions
> > on
> > the best spam filtering software? I've tried spam assassin, but even with
> > tuning, and a few GB of spam to train on, its results are pretty abysmal.
> >
> > Thanks,
> > Jordan
> > _______________________________________________
> > Discuss mailing list
> > Discuss at bozemanlug.org
> > http://lists.bozemanlug.org/mailman/listinfo/discuss
> >
> 
> _______________________________________________
> Discuss mailing list
> Discuss at bozemanlug.org
> http://lists.bozemanlug.org/mailman/listinfo/discuss