----- Original Message ----- 
From: "R. Scott Perry" <[EMAIL PROTECTED]>

> The reason for that is that it isn't easy administering a DNSBL -- there
> are a lot of costs involved.  We're expecting to get tens of millions of
> queries per day.  If someone else was running this test, that would be a
> different story (for example, if we limited the SPAMCOP test to those with
> a Service Agreement, I think that would be wrong).  Plus, this is
something
> that isn't available anywhere else.  It is essentially a new feature to
> Declude -- and as such, should require a Service Agreement.  It is a
> service, and as such really shouldn't be free.  I think that it should be
> seen as an extra benefit to the Service Agreement, making the Service
> Agreement more valuable.  FWIW, I do agree with management that this
should
> only be available to those with an active Service Agreement.

And how are you preventing anyone but current customers with active SAs from
using the DNSBL?  If someone knows the test site hostname, what is to
prevent them from using it?

Also, it does not appear that this IP4R test is very robust as almost all
queries posted to the server fail with a response of "srvfail" or timeout
with "no servers could be reached".  Right now it is very rare for
mtldb.declude.com to come back with a valid response, either positive or
negative.

I've got to say that Computerized Horizons struck-out big time on this one.
Not only is the test flawed (hitting way more legit mail than spam), it was
supplied in a most bizarre fashion.  One of the most appealing thing I found
about the Declude products was the fact that they were so open and
understandable, much like open-source software in the UNIX/Linux world.
However, this last update changed all of that and, thus,  left a lot to be
desired.

I know that change is inevitable, but I don't recall anyone asking for
graphical installs or graphical upgrades on this list, and I for one hope
that Computerized Horizons goes back to the old tried-and-true methods that
have worked so well in the past.

In the mean time, I have disabled this test because of it very high
false-positive hit rate and extremely low valid hit rate on actual spam
message.

Bill

---
[This E-mail was scanned for viruses by Declude Virus (http://www.declude.com)]

---
This E-mail came from the Declude.JunkMail mailing list.  To
unsubscribe, just send an E-mail to [EMAIL PROTECTED], and
type "unsubscribe Declude.JunkMail".  The archives can be found
at http://www.mail-archive.com.

Reply via email to