In 1991, I met Joe Judge, who was the postmaster for one of the biggest and most respected domains on the Internet, att.com. At the time, I was with the Explorer Post 891 (think: career-oriented scouting and you pretty much get the picture), sponsored by AT&T; Network Systems and Bell Laboratories. The post advisor managed to attract Joe's interest and he, along with Danny Zerkel and Beau Smith, introduced us to the weird and wonderful world of lpMud.

LpMud was a multi-user role-playing game, where players could work together to go on quests, working their way through Middle Earth (and who knows what else), slaying orcs and solving puzzles on their way to their objective. The idea of introducing lpMud to the explorers in the program was that it would be an excellent way to hold everyone's attention while also giving them the ability to program a real system that other people would use, in a sort of simple quasi object-oriented language called lpC. (Plenty has been written about lpMud and lpC already; there is no need to recount the whole history here.)

I got to know Joe while working with him on the Explorer Mud, hacking the system so there would be a game for people to play, and trying to coordinate the efforts of other explorers who were working on the project. The whole lot of us spent a fair bit of time working on the Explorer Mud. Some of us turned out to be fascinated with the development of the system to support a programmatic environment, others were drawn to the development of the game and the puzzles for players to solve, while yet another group liked mostly to play the games once they were built. As is always the case, we also had our own collection of socialites who didn't serve any obvious purpose for the development of the game but in fact played the important role of keeping the budding "community" together (and occasionally at one anothers' throats).

As I spent more time with Joe, I started to see what his job was all about at AT&T.; His group was responsible for connecting all of AT&T; worldwide to the Internet, and doing so safely. In addition to dealing with the enforcement of network policy (at a time when there was no such thing as a firewall product that one could buy), the group needed also to provide connectivity between different types of networks, e.g., AT&T;'s proprietary DATAKIT network (used extensively inside of the company) and the IP-based Internet. Joe dealt with it all: operating systems, applications, networks, and needed to understand how all of those things worked. It wasn't just being a generalist in the sense of not specializing in anything: he had to have a specialist's understanding of all of those things all at once. I remember thinking that his job was probably the coolest in the world, and if I could work in the same group that he did, I would be set.

Joe introduced me to other newfangled technology in the Internet world. By 1992 or so, I was well familiar with basic services like Internet email and Usenet, to the point that I was looking at how to implement software that would handle these better than the other options of the day. In 1993, Joe introduced me to a piece of software called NCSA Mosaic, the first graphical "browser" for something called the World Wide Web -- which basically served the same purpose as Gopher, except that it could be multimedia and hypertext. Later that year, I purchased a Sun SPARCstation 1 (Sun 4/60), connected to the Internet on a 14.4 kbps modem running the CSLIP protocol, and ran Mosaic from home. Because I was running SunOS 4.1.3, I did not have the libraries to resolve domain names (like www.cis.ohio-state.edu) to IP addresses (like 164.107.123.6). That was fine for most of my purposes: I knew the IP addresses of all of the systems I would ever connect to. The Web changed all of that, linking things from all over the place, potentially from systems that I never had heard of. A few hacks to the operating system later, I had DNS resolver libraries and I could click all around the Web.

At the time, I had a list of all known Web sites (there were only a few dozen) and I started to collect technical documentation on things like the HyperText Markup Language (HTML) and the HyperText Transfer Protocol (HTTP) that made all of these things work together so seamlessly. Not long thereafter, I would have a strange job in a financial services company where I wore the hats of systems programmer and data processing supervisor and would find ways to use Web technology to solve serious problems that were facing us in our ability to look at data the way that we wanted to see it. Some Perl programs written to use the Common Gateway Interface (CGI) to the NCSA Web server gave us lots of options that we didn't have before.

Fast-forward a few years, and Joe had left AT&T; (and Ohio) for the financial services industry (in Boston). I managed to land a contract job not just in Joe's old group, and left the financial services company I worked for. In case that switch wasn't strange enough, I was actually responsible for much of the job that Joe had when he was at AT&T.; I started out as a system administrator, dealing with things like the postmaster email, basic systems operations, making sure that things would continue to work for the tremendous load of email and Web traffic that flowed between the Internet and the AT&T; internal networks.

The AT&T; corporate Web site officially went live in October of 1994. My group there was responsible for the technical operation of that site. Under the technical guidance of Gary Ellison and with expert technical assistance of my friend Brian Larkins (who changed to our group not long after killing AT&T; System V machines in another group by invoking the Perl interpreter), AT&T; had a real Web site. (Of course, there were a few other minor players like me in Columbus who dealt with the Web site from time to time, and some major players at various points back in New Jersey. I was focused on the firewall architecture and dealing with proxitized Internet access for many different protocols besides the Web, as well as a relatively complex architecture to support email, so I didn't work intimately with the New Jersey contingent of our group much of the time.)

Now, here it is, late September 2004. Not until I got a copy of the ten-year anniversary email from the old AT&T; posse did I realize how long it had been, or to spend much time reflecting on all that had happened. I clearly remember the late nights in the lab at Part 17, with Brian shouting, "IT'S TIME TO DO THE NEWFS!" but somehow it seems distant now.

Within a year of that Web site going live, AT&T; would decide to split off NCR (which it had forcibly acquired in a hostile takeover not long before) and to form a new spinoff company called Lucent Technologies. Many of the old posse would go off into different directions. Gary to Sun, and I first to an advertising agency that was doing work with integrating the Web into the advertising that its clients were doing in print and broadcast media, then to an Internet startup that ultimately had a lot of the technical folks from the New Jersey contingent of the old AT&T; Web group. Gary even did some crypto work for us after leaving AT&T; and before actually moving out to Californ-i-a to start at Sun. I'd be rich if it worked, but in an action to which I have since become accustomed, the company I worked for led the way for the industry, and we tanked in 1997 -- on the same week that Computerworld's "Superprogrammers" article declared to the world that I couldn't be hired by anyone for any price. (That did, in fact, complicate my subsequent job search somewhat.)

And now I work for a company that came out of the research efforts that some of the old posse members, notably Gary, Doug Monroe and I, were making into how systems that people were using impacted their privacy and security. We have a few papers on our findings, a book on building Web-based applications and sites that respect user privacy and data security, and the work has been cited in lots of different places, both formally and informally.

Funny that ten years had to go by before I started to put things up on the Web that were outside of the scope of my professional work. I guess it might said that like thousands of others, I've been so busy building the Web that it wasn't until recently that I really started to use the Web. Somehow, it seems that the most appropriate comment on my work of the past decade is the simple fact that I'm now using it in my daily, non-work life. What started out as a strange technical curiosity to be understood and played with has become something that affects my own life. What a long, strange trip it's been.