On Sun, 1 Jun 2003, Anthony Atkielski wrote: > David writes: > > > Guessing and trying? > > That would require tens of thousands or even millions of bounces for every > successful mailing attempt. I don't think anyone is doing it that way. Paul Vixie has provided his specific proof. I'd suggest that a very high percentage of 3 letters combined with aol.com, msn.com, yahoo.com and earthlink.com were/would be found this way. Furthermore, all common first names and last names and combinations. Two letters with common last names, etc. > > > Online directory provided by her email provider? > > Which e-mail providers are providing online directories? And how many of > these directories allow you to extract addresses without specifying any > search parameters (such as a name)? When I found AOL's member directory, a simple search on my last name was sufficient to obtain a result. Commonly available windows QA tools would allow automation of the AOL GUI to harvest a large list. Needless to say, I closed the barn door, probably too late. The other responses were based on differing evaluations of human behavior. Since neither of us have documented credentials in that space my comments were speculative as were the responses. I don't think many conclusions about address harvesting can be drawn from personal experience of folks who are users of small domains or technically savy users who avoid risky behaviors. Dave Morris