[Info-vax] Anything wrong with this anti-spam measure?

[Phillip Helbig---remove CLOTHES to reply]

Running an SMTP server at home, I get a lot of SMTP connections.  Most
are spam.  I get about one connection per second.  Using the
spamhaus.org RBL and setting Symbiont-Checks-Deliverability: to FALSE in
SMTP.CONFIG gets rid of most of it before it even gets to VMS MAIL.  
Most of the stuff is from addresses in the RBL or to a non-existent user 
or both.

Unfortunately, I am still at TCPIP V5.4, so an email to a non-existent
user which is NOT a syntactically correct username gets through
(assuming it doesn't come from an RBL address).  I hope to upgrade VMS 
and TCPIP by the end of the year.  IIRC, this has been fixed in a newer 
TCPIP.  Right?

Until recently, this was not a problem, but for a few weeks now I have 
been getting A LOT of spam to three addresses.  This is a pain, and the 
rejection messages probably generate backscatter spam.  Thus, I did the 
following (IIRC, this was originally suggested here by JF): create an 
empty file called NL.DIS in TCPIP$SMTP_COMMON, then do this:

   MAIL> SET FORWARD/USER=<non-existent user> SMTP%"NL at multivax.de"

Although the mail does get as far as VMS MAIL and thus consumes more 
resources, otherwise it seems to work OK, i.e. it disappears silently 
and has no side effects.  Can anyone think of any potential problems 
with this approach?

(By the way, I recommend defining TCPIP$SMTP_COMMON to be a search list
containing the default SYS$SPECIFIC:[TCPIP$SMTP] as well as a
user-defined directory for use with all nodes in the cluster.  .DIS 
files and so on can go into this common directory.)