[Mimedefang] Re: compare mimedefang to mailscanner

[John Rudd]

David F. Skoll wrote:

> I'm not trying to say that MIMEDefang
> is better or worse than Mailscanner.  I was just trying to objectively
> (as much as possible!) compare their behaviours under different
> load conditions.


And I would like to second that statement.  I've used both quite a bit 
in both home environments and in mission critical production 
environments.  I don't think MailScanner is an inferior package.  With 
one exception (the order in which it does its checks) it is VERY good at 
attacking the anti-virus/anti-spam problem with the strategy it uses. 
It's just not the strategy I prefer.

MIMEDefang is _also_ VERY good at attacking that problem with its own 
strategy.  It's just not the same strategy that MailScanner uses.  And, 
clearly, I prefer MIMEDefang's strategy.


The one thing I would say is a weakness with MIMEDefang is that, as 
Scott suggests, there's a little bit of a learning curve if you want to 
strike out on your own.  I've been writing a mimedefang-filter that has 
an extensive set of "on/off switches" and "config variables" at the 
start.  With input from multiple people, that might make an interesting 
way to soften the learning curve (and make it all more usable for 
non-experts in trying to switch on and off different behaviors).

But the other side of that coin is: if you need that kind of 
hand-holding, you might be better off paying for canit-pro.  It uses 
MIMEDefang at its core (right? I didn't misinterpret that?), and wraps 
around that a support/etc. package.  I'm willing to bet that it ends up 
doing a lot of that kind of "softened learning curve" stuff for you.