On Mon, 7 Jul 1997, Ken Nakata wrote:

> There's a port-mac68k subscriber with address ssartori@cnj.digex.net,
> but there's noone with address ssartori@cnj.digex.net.digex.net (not
> surprisingly!).

Shell of /usr/local/bin/cancel... hmm... last login May 13?  Sounds like a
dead account to me....  Dunno.

<snip>

> And last thing but not the least you should consider is the ethics of
> forging an email message, even though it is pretty clear that the
> subscriber is no longer getting any message from us.  Besides, you
> might be held responsible for doing or even suggesting such an act,
> David.  I'm not sure if it violates any laws, though.

Which is pretty much why I tried to make it clear in the first of the
message that I considered it a Bad Idea (tm).



On a related topic... what's the best way to make sure that bounce
messages go back to the original sender?  Some mail servers seem to
totally ignore the contents of the document, i.e. From: and Errors-{T/t}o:
fields.  Moreover, sendmail (at least under MkLinux) sends DNS errors
to... whatever user runs the program that opens the pipe to sendmail. 
Unfortunately, the program I'm working on has to run setuid root to access
all the files it needs to access.  Thus all that junk comes back to me at
the root account instead of to the talker-errors account.  What's the best
solution to that?

And does anyone know if there's some fairly simple set of rules to send
select messages to ds instead of talker-errors if they involve any form of
nameserver timeout or mail server error other than an invalid user or a
truly invalid host?  I'd like to filter legitimate bounces and send them
to talker-errors, but send things caused by screwed up mail sysems to
either ds or root or something.  Ideas?


Anybody?
David

-----BEGIN GEEK CODE BLOCK-----
Version: 3.1
GCS/CC/FA/H/L/MC/M/MU/PA/TW d-@ s:>- a-- C++++>$ UBLAS*++++>$
P+?>$ L+++>$ !E--- W+++>$ N++(+++)>+++$ !o? K-? !w--- !O
M++>$ !V-- PS+>$ !PE- Y+>$ PGP+>$ t+++>$ 5+>++++$ !X- !R tv+>$
b++>$ !DI !D- G++(+++)>$ e>++++ h--! r--- !y-
------END GEEK CODE BLOCK------