Commercial filtering (was Re: [Mimedefang] ADMINISTRIVIA:Out-of-office replies)

[Cormack, Ken]

> Too bad. :-(  What's the other product?  Any chance you could evaluate
> CanIt? :-)

The product is Secure Computing's "IronMail".  They've had two filtering
servers in service for the last several years.  They've just installed the
Centralized Management console, and Centralized Quarantine server.  Our
filter servers are being racked next weekend, and the vendor will be here to
work with me the following week, to get them configured and talking to the
central servers ar the other campus.

> We're not even on Gartner's radar, obviously.

I tried early on, to argue the success rates we'd had with MIMEDefang and
Sendmail, and the vitues of having complete control over the filters via
it's open source nature.  But the partner campus (main headquarters of the
company that bought us) has little historical experience with Open Source,
little real creative "hands on" experience developing any tools (I'd be
shocked if anyone there had ever written a shell script bigger than a few K
in size), and in general, the philosophy there had always been to go with
commercial products.

I tried to pitch CanIt for consideration when their initial (and very lame)
counter-arguments against MIMEDefang were limited to "it doesn't have a
GUI", and "there's no vendor support".

Paying for vendor 4-hour response time doesn't mean the problem is solved in
4 hours... It means the vendor is only contractually obligated to call you
back within 4 hours.  I gave them evidence of the very quick responses and
excellent suggestions I received to questions posted on this list, including
several rapid responses directly from the primary developer (you).  They
were'nt impressed.

This attitude, I think, got very deeply ingrained in them back in the early
days of the SCO lawsuit threat, and they were afraid of potential liabilites
assigned to any "Open Source"/GPL-type stuff.  And that was right at the
time they bought us.

At my own campus, had used a lot of open source solutions (BIND for DNS
instead of Lucent's QIP, MIMEDefang and Sendmail instead of IronMail, Linux
instead of Solaris, etc.)  Over the years I had invested a lot of ::ahem::
"passion" to defend the Open Source solutions, but I ultimately lost each
battle.  Technical and Business arguments (ROI, Initial/ongoing costs) were
tried, but with no success.  Again, I think it's a personal victory that the
solutions have remained as long as they have, but having taken a beating for
my persistance in 2 past performance reviews, I decided I'd best just be a
"Yea, Team!" Player and accept the solutions being brought in.

Ken