My publications are available from the University of Kent's Academic Repository.
Recent talks include:
- Provenance Tracking in R, presented at the International Provenance and Annotation Workshop (IPAW2012), Santa Barbara, California (June 2012);
- CXXR: An Ideas Hatchery for Future R Development, presented at the Joint Statistical Meeting (JSM2011), Miami Beach, Florida (August 2011);
- Interpreter Internals: Unearthing Buried Treasure with CXXR presented at the useR!2011 conference, University of Warwick (August 2011).
I belong to the following research groups:
Since retiring from a full-time academic post, my chief area of research interest is in statistical computing, and particularly the CXXR project to refactorise the R interpreter into C++.
I wrote my first computer program in 1960, for a first-generation machine called the Ferranti Pegasus installed at The Steel Company of Wales at Port Talbot, where my brother worked at the time. The 'fast' memory of this machine, corresponding to RAM nowadays, comprised 56 words of 40 bits (i.e. 280 bytes), implemented using nickel delay lines. The backing store was a magnetic drum holding about 25 kB.
Between school, where I specialised in maths, and university I worked for nine months at the Royal Aircraft Establishment at Farnborough, where I came across another Pegasus machine, by that time not much used, and it was on this computer that I cut my teeth in assembly language programming, culminating in a program to calculate pi to 1000 decimal places. I ran this one lunchtime: it took an hour to calculate the answer, and then ten minutes to convert it into decimal and punch it out onto paper tape.
My first degree was in Philosophy at Cambridge, then called Moral Sciences, specialising in logic; during long vacations I returned to RAE Farnborough where I worked in the systems programming group. After that I proceeded to a doctorate, again in the Faculty of Philosophy, on the logical foundations of statistics and predictive reasoning.
After ten years in philosophy, I found I wanted to do something more practical, so I joined the inertial navigation division of a large avionics company in Rochester, Kent, which kept on changing its name, but by the time I left was called GEC Avionics. (It is now part of BAE Systems.) There I played a key part in the company's development of terrain-referenced navigation systems, and was the inventor of a system called SPARTAN, which was flight-trialled on Tornado, A6 and F16 aircraft, and went into full-scale development for the Tornado; sadly, it never entered production. My work in this field led to my being elected as a Fellow of the Royal Institute of Navigation.
As time went on, I rose up the hierarchy, eventually becoming 'Manager of Systems, Software and Future Projects', and life became full of writing proposals for new contracts, and fire-brigade actions in support of current projects, with little time for the technical work that I enjoyed. I eventually had enough of this, and came to the University of Kent in 1988.
Since I am only (very) part time, there is usually not much point coming to my office on spec. Instead, if you want to contact me:
Email: I normally read my university email several times a day. Please use plain text in any email to me if at all possible.
Telephone: In cases of at least moderate urgency I can be contacted by telephone at home on 01795-535697.