One of the ways that spammers get valid email addresses is from websites where email contact information is posted. Some folks attempt to disguise their email addresses — so that humans can figure them out but programs can’t. Of course, humans continually enhanced programs to make them better — so most of those techniques only delay the harvesting.
I’ve been thinking about an alternate approach. Would it be possible for the server to only display a valid email address to a human, not a robot? In other words, could the server be smart enough to detect the kind of client requesting the web page and dynamically adjust the content of the page accordingly?
For example, could I include a small piece of Javascript code in my ‘protected’ page which would interrogate certain elements of the client’s browser — beginning first with a set of tests that attempt to ferret out whether the client is likely to be a human vs. a robot? Then the content would be dynamically adjusted to either provide the client with a valid email address — or — a spam address.
Oh, by the way, consider this an invention disclosure — it is now prior art for anyone else attempting to patent this particular approach.
Update: I’m getting some great comments. Comments that the approach I suggested won’t work — because spoofing, if it becomes popular, will be cracked. Another solution suggested is to put your email address information into a graphic — which can’t be harvested/decoded by a spammer.