Published in the Dallas/Fort Worth Unix Users Group Newsletter, July 1997.
As an avid Usenet participant and a reluctant email mailing list administrator, I thought it would be interesting to discuss a trend that I've seen lately in these two worlds. Granted, this is only my views based on one particular newsgroup (comp.software.testing) and one mailing list (swtest-discuss), but perhaps the same phenomenon occurs elsewhere as well.
The phenomenon that I refer to is that some people are abandoning Usenet and choosing to use mailing lists instead. While the charter of the swtest-discuss mailing list isn't exactly the same as comp.software.testing, for some people it's close enough, and they seem to be using the mailing list as a complete replacement for the newsgroup. It may be because the signal to noise ratio is getting worse as the number of Usenet participants skyrockets. Or it may be that anyone unfortunate enough to ever post anything on Usenet can expect to receive a continuous stream of junk email for years to come.
These problems are often not as bad on mailing lists. I control the mailing list from a central place. This isn't the case for Usenet newsgroups unless they're moderated. I have chosen not to moderate swtest-discuss (i.e., read every message and decide whether it gets posted or not), but using the freeware majordomo mailing list software, I can tweak several configuration options to set up various levels of control. For example, I disabled the command that allows people to retrieve the list of members on the mailing list. Without this option, a mailing list is actually less private than Usenet, since no one can tell who's lurking on Usenet. Since I archive all messages to the list, people can find out the addresses of people who have sent at least one message to the list, which is the same sort of problem that causes people who post to Usenet to get their addresses on bulk email lists. However, it is somewhat more restrictive on swtest-discuss because you must be a member of the list in order to retrieve information from the archives. I suspect that many of the scumbag spammers who dredge Usenet for email addresses would be unwilling to bother to subscribe to the mailing list in order to get a few more addresses, because they'd leave a public record that they had been there, and they'd have to subscribe using a valid address in order to be able to retrieve the information.
Regarding the signal-to-noise ratio, this also tends to be much better on the mailing list. The charter is fairly specific, and it is enforceable because I have central control over the list. On a few occasions, I have chastised list members for posting messages that were not allowed by the charter. They either fell into line quickly, or voluntarily left. I haven't even had to forcefully remove anyone for violating the charter. Another factor is the fact that only list members are allowed to post to the list. This means that everyone who sends a message has already received a copy of the charter and has no excuse for violating it.
It's interesting to note that some Usenet newsgroup have been trying to shoehorn some of these measures into the Usenet architecture. One idea is to make a newsgroup moderated, but instead of giving some poor soul the burden of moderating the group, you automate it somewhat. For example, see the recent comp.lang.perl.* RFD at ftp://ftp.uu.net/usenet/news.announce.newgroups/comp/comp.lang.perl-reorg2. A robot script will automatically filter out some postings, and it will maintain a list of approved posters. It adds new posters to the list after sending them some introductory information. A team of moderators oversees the robot, and they presumably can manually change the list of allowed posters. So some of the benefits of mailing lists are slowly becoming available on Usenet, though there may never be as much privacy on Usenet as you can have on a mailing list. An even lighter-weight solution is to use a robot on an unmoderated group to send the group's charter to all new posters. This was discussed in comp.software.testing, but we decided to frequently post a brief pointer to the charter instead.
I said that I'm an avid Usenet participant, but only a reluctant mailing list administrator. One reason for this is efficiency. It would take some research to determine exactly how the efficiency of Usenet and mailing lists compare, but I'm confident that Usenet is a more efficient delivery mechanism to reach large groups of people. There are only about 200 people on the swtest-discuss mailing list, though, and it might actually be more efficient to use the mailing list to reach 200 people than to post a message that goes out to every news server in the world. But what if all of the estimated 18,000 comp.software.testing readers abandoned Usenet and joined swtest-discuss? The kind folks in my office who provide the host machine for the mailing list wouldn't be nearly so kind if that happened!
Mailing lists seem to have an exclusionist air about them. The relatively small group of people on swtest-discuss seems to enjoy the fact that not many people have joined the list. This was even more the case when the group was smaller and the charter was much more focused. A less sinister explanation is simply that people like to hang out with a smaller, more intimate crowd. In fact, people tend to open up more when they know that fewer people are listening. I think that small mailing lists will continue to fill a need for small, focused groups, because the Usenet community frowns on creating newsgroups that have a very small readership, and because mailing lists have more privacy. For larger groups, though, and even medium-sized groups like swtest-discuss, I'm hoping that the emerging techniques for improving the quality of Usenet newsgroups will help to reverse the migration away from Usenet. Then if we could fend off the email spammers, Usenet could regain some of its former glory.
Copyright 1997 by Danny Faught