From 1988 to 1992 I worked in the friendly computer group at the CSIRO Division of Radiophysics and the Australia Telescope National Facility. I was allowed to dedicate one-third of my time to astronomy, the rest to managing the Convex supercomputer and a fledgling Sun Unix workstation network. I also had two small children, so that went about as well as you might imagine.
In mid-1992 I left astronomy and moved back to my home town, Newcastle, and took a job as system manager of the Sun workstation network for the Department of Computer Science, University of Newcastle. I liked my colleagues and grew to enjoy the work — the Internet's momentum was gathering pace and it had its own fascinating problems, both human and technical.
Crafting a Code of Ethical Conduct
Systems Administrators Guild of Australia Annual Conference, Perth, 1994.
Managing computer systems in those days was a lonely business and not for the faint-hearted, but the Systems Administrators Guild of Australia started up in 1993 to try to understand our roles in the new globally-connected Internet. SAGE-AU wanted a code of ethics and I ended up herding the cats of the committee, some of whom were vocal in their belief that the Internet was but a passing fad. Here is a code of conduct written in more innocent times.
Network Monitoring Tools
Systems Administrators Guild of Australia Annual Conference, Wollongong, 1995.
There were few commercial monitoring packages to quickly recognise and deal with network problems, so mostly you rolled your own from open-source utilities. Also, and incidentally, since this conference I have never again been able to drink gin and tonic.
The Nature of Network Traffic
QUESTnet, Brisbane, 1995.
Traffic on computer networks behaves oddly. I would have delivered this paper at QUESTnet, but I booked the wrong flights and missed the conference. So the papers is in the proceedings but no one ever witnessed it.
Introduction to Unix
Department of Computer Science, The University of Newcastle, 1996.
A Unix userguide: there are many such in the world today, but there were few then. I left Newcastle in March 1996, lured away by a job in Melbourne with Connect (connect.com.au), the first commercial public-access Internet service provider in Australia. I managed the talented system group, running the Unix computers that provided Australia-wide access and email services, and later also managed the programmer group.
Systems Administrators Guild of Australia Annual Conference, Melbourne, 1997.
How to find good sampling intervals for monitoring the long-term performance of machines and see when resources are degrading to potential bottlenecks. Yes, demand has always outstripped supply.
Quantifying Chaos: Metrics and the Internet
Australian Unix Systems User Group Annual Conference, Brisbane, 1997.
To me the most interesting problem on the Internet at that time (others may disagree) was the way traffic behaved in a self-similar manner: it was like a fractal, showing the same structure at every level of resolution.
The Domain Name System: Engineering vs Economics
Australian Unix Systems User Group Annual Conference, Sydney, 1998.
This was about the founding of the Internet Society of Australia, the transition of a community to an industry, and how a vital and beautiful piece of Internet engineering became the morass of exploitation it is today.
Who is an Internet Content Host or an Internet Service
Discussion Paper, Internet Society of Australia, 1999.
A calm, sensible attempt to explain to the right-wing government of the day it didn't have the faintest idea how the Internet worked. Even based upon its own definitions, the laws it was busily implementing had big ugly implications. Our words were ignored. We didn't understand it was all about the press releases.
Connect was taken over, and in April 2000 the new management decided I was surplus to requirements: apparently a very nice guy in a suit replaced me. It was painful but also a stroke of good fortune, because later that year I ran across the story of the mysterious lugger Redbill, and found out I liked researching and writing history much more than dealing with business spivs.