My USENET postings have quite odd addresses in them. This is the explanation. See my FAQ for more information.
From: Richard Kettlewell <richardk+ah-from=oxbridge.tat@chiark.greenend.org.uk>
Newsgroups: oxbridge.tat
Subject: Re: Happy New Beer! (Was: Re: Moments of Sheerness)
Date: 02 Feb 2000 16:21:23 +0000
Robin Stevens <robin.stevens+randomword@oucs.ox.ac.uk> writes:
> Btw, what's "ah-from"?
A little while ago I decided to track where people (and other entities) were getting my email address from.
So, I started putting the newsgroup name in the From and Reply-To header fields on my postings. The reason for using both and having different values for each is to see where the contents of the two headers end up.
For example there is a theory that address harvesters pick up the From field as it's much quicker than picking up the Reply-To field, since the From field appears in news server overview data, while the Reply-To field does not.
Another interesting place that these addresses end up is the attributions added by news client software when generating follow-up messages; some of them use the From field and some use the Reply-To field. (Personally I think they ought to use the From field.)
Anyway, recently I spotted the first example of a piece of spam addressed to richardk+from=<something>. SAUCE correctly rejected it, as it happens, but I decided I wanted to change the address so I'd be able to distinguish future instances of address-harvesting from the one that led to that address.
The extra thing I decided to put in the addresses was members of the set of two-letter words in the nearest available word-list (not necessarily in any particular order). Next time I spot a rejected spam, I'll change it again.
Even if I change once a month, I have about four years before I have to switch to a new scheme.
HTH