© Copyright 2000 by John Halleck
[Still under construction]
Various computer systems handle email addresses in different manners.
In the early days of computing they were they account name. Nowadays they are usually driven out of a table. Many sites use things like Joe.Blowtowski@foo.bar, while some use things that are more cryptic such as firstname.lastname@example.org. Some sites even do both, with higher ups getting their names and peons getting the cryptic stuff.
The choice of what sort of addresses a site uses are determined by the administrators. The fancier addresses are more work to install and more work to keep current. For example, they have to be updated every time someone's name changes. For a university, where there are lots of people of an age likely to marry, this can happen a lot.
Normally one would consider the choice truely an administrative one, but there can be some ethical side effects of the choices.
Multiple people can have the same name. (This university, for example, at one time had an EE class that had THREE people with exactly the same name.) This brings up the issues of who gets the name, and causes problems for the person that doesn't get the name.
People out on the net note that a site has, for example, names of the form Firstname.Lastname@foo.bar. Therefore if they know their friend's name, they will assume that the friend has an address built in that manner. Given the unbelievable tendency of people to send amazingly personal stuff to an address that they don't know works, this means that the wrong person (with that name) learns all kinds of interesting things about the intended person. They also inherit the stalkers, bill collectors, and other fun baggage of that other person.
This same sort of problem exists with telephone books, but if someone calls up the wrong person they rapidly notice (wrong voice) and the process doesn't continue to the point of mistakenly sharing intimacies. With computers, on the other hand, they progress to the point of mistakenly sharing information before any feedback is received.
Cryptic names that one sees as email addresses are usually generated by some mechanical algorithm. For example, a person's initials plus a number to make it unique is popular. So the first person with the initials JB gets JB1 and the second gets JB2, etc.. This scheme makes it easy to generate the ID's for students before they even show up if you have the records.
Unfortunately, this means that addresses that are "close" to a given address don't fail, they just go to someone else. So if I mis-type (or mis-read) the address jb7 as jb2, I don't get an error when I send, and someone else gets the mail. If the address is part of a CC (carbon copy) list, I might not even notice a problem since the person that it is really sent to would reply, and I wouldn't expect people CC'd to reply.
The situation in the previous paragraph may sound artifical, but it actually happens with more frequency than one might expect. We had a case at this university the police getting involved in a problem with arguablly illegal mail, because of exactly this issue. The sender was sending to a student at this university, and was CC'ing what he thought was the student's account here. But in reality the address was that of a faculty member that was on leave. When they FINALLY got around to reading their mail they discovered that they had a large collection (the complete history) of what was going on. The faculty member brought this to the attention of their local administrators, and the administrators had little choice but to turn it over to the police. It became a rather ugly scene all around.
While the problem above was something that involved the police, much subtler privacy issues are not hard to imagine. Everyone from lovers to businessmen tend to have information in their letters that would at least make them feel bad if released to some random individuals.
There really is no "correct" choice.
Some have suggested letting people pick their own email name, but this doesn't really solve the problem either. "I"ve used that nickname on IRC for years, but it already assigned to someone else." The collision problem with "self chosen" names tends to be greater than with people's actual names, and the name collision issues are actually worse. (For example "Mycroft" is almost unknown in real names, but is commonly picked by people for themselves.)
One could chose randomly generated email addresses <email@example.com>, chosen by program to not be "close" to the spelling of any other name, but those are practically impossible for most people to remember and it makes it horrendiously difficult for people elsewhere to find, much less type, the address of the user here.
[UNDER CONSTRUCTION] Please feel free to coorespond with me on this issue.
This page is http://www.cc.utah.edu/~nahaj/ethics/addresses.html This page is © Copyright 2001 by John Halleck This page was last modified on January 20th, 2001