This article at the New York Times made me laugh quite a bit this morning:
LAST fall, David A. Green was looking forward to his new job at a Manhattan real estate brokerage. Finally, he would be able to shed the e-mail confusion that plagued him because of the two other David Greens at his company.
No such luck. At his new company, Cushman & Wakefield, there was yet another David Green. And that Mr. Green already held the coveted e-mail name David_Green, following the company's convention for e-mail addresses.
At his old company, CB Richard Ellis, he had been, reluctantly, David.A.Green. Now he became David_A_Green. Hundreds of his e-mail messages went to the wrong David Green.
"I would have been better off being at his desk than at mine," said David A. Green, noting that the original, initial-free David Green is "most gracious about sending me the volumes of e-mails that go to him." [via New York Times: Technology]
So why did I find this amusing? This isn't a new topic. The sendmail FAQ has had a question on this forever (or at least since 1997). It still amazes me that people don't think that this might be a problem. When I was in college pretty much everyone had four character email addresses in the format (first initial)(second initial)(random alphanumeric)(random alphanumeric). Mine was gb1d. I can still remember pretty much all my friend's addresses from then too. It also helps to keep out people just spamming by sending emails to any old firstname_lastname address.