[antispam-f] Re: Recent Spam

  • From: Jeremy Nicoll - freelists <jn.flists.73@xxxxxxxxxxxxxxxxxxxx>
  • To: antispam@xxxxxxxxxxxxx
  • Date: Thu, 29 Nov 2007 18:19:18 +0000

Dave Barnett <as10@xxxxxxxxxxxxxxxxx> wrote:

> I'll add this to my file.  As Jeremy Nicoll, in message 
> <gemini.js8j1l00300ho00ds@xxxxxxxxxxxxxxxxxxxxxxxxxxxx>, points out, 
> there are risks with this test, 

...yes

> especially if you have correspondence from ex-colonial countries located
> West of time zone +0500. 

What?  How can that be related to the format of an email address?

> Using this on the To: header shouldn't be a problem, as I don't use the
> format in any of my names.

But it's perfectly legitimate that you receive an email that's eg:

  To: someone-else
  BCC: you

in which case there will be nothing in the To: line (and nothing necessarily
in any other header) that contains your real name.  Many mail lists work
that way, which may be ok if they're all actively looked for by ACCEPT
rules.

Also if I sent a mail eg to another AntiSpam user, but thought it might
interest you, and didn't want them to know I'd copied it to you, the TO:
value might have their name (in any format) and only the SMTP enevelope info
would actually route the mail to you.


> The combination should be as foolproof as anything is when dealing with
> spam.  I'll start, as usual, with a Header trial.

There's no doubt it will delete things, but that's not a reliable test of
whether it deletes the right things.  It may delete lots of garbage, but
then also delete genuine mail to you.  How would you know?

-- 
Jeremy Nicoll - my opinions are my own


Other related posts: