In article <7f26t1-(E-Mail Removed)>, Doug Laidlaw wrote:
>Some spammers are getting past it by using my address as the "From"
>address. I may talk to myself, but I don't write to myself - yet.
Hi Doug,
On the systems at work, I always Bcc myself, so that does show up as
mail from me. However, the mail filter knows to test first for the
presense of a less obvious header, and whitelist my mail on that.
On the rare occassion when I mail something from the home accounts
to a work address, I've learned to be using a non-publicised address
that changes regularly which is created using a hash of the output of
/dev/urandom.
>Is there any harm in putting my own address in a blacklist?
If you _never_ send mail to yourself, then no. If you do send mail
to yourself, look at the headers of such good mail, and you'll soon
notice _some_ header that is unlikely to be forged. For examples,
this could be a header you've added (sorry, I don't use KMail), or it
might be a specific series of "Received: from
dialup-1-118.Bendi.mumble.Mumble.com" headers. In any case, you can
whitelist on that, and not worry about blocking based on anything else.
Do remember that there is a difference between the "envelope from"
address (Return Path

, and the internal (to the mail) From: address.
Both are easy to forge, but the internal address is childs play.
Should you be running cron jobs that might create mail froom you to you,
(usually errors, but occassionally I may want cron to mail me for some
reason), again look at known good mail. Often, the internal from header
is from
(E-Mail Removed), rather than
(E-Mail Removed).
Like my old Unix instructor said: "know your files" - meaning know what
the stuff you'll be looking for looks like, as well as the rest of the
crap that you don't want.
Old guy