Thesis Overview

---

Chronology

The thesis started in May 2001 and went as follows:-

• May was spent primarily researching the characteristics of mobile phones.
• June was spent going over the various algorithms, working out what needed to be done and making sure that the various parts would fit together.
• July was spent actually coding and arranging for a server to be installed
• August was spent installing the various programs on the server.

What went right?

• The thesis actually works. The brief was to develop a hand-held device that would guide you to your destination and that’s precisely what has been developed.
• The programs are small, easily comprehensible and carefully defined, which is what a program should be.
• The programs run efficiently
• The programs are expandable to include things like new human languages ( French, Japanese, etc) and new machine languages (cHTML being the obvious candidate).
• The introduction of birth and death days has enabled the system to deal with jouneys by road, train, plane, ship, foot, or indeed any combination thereof.

What went wrong?

• The Herald account (the original choice of server) proved to be inadequate when it became apparent that the University’s safety restrictions would hamstring such a program. This was initially a serious problem and moves were made to get a commercial hosting company to host the programs instead. Thankfully the University support staff stepped in and saved the day by installing a brand-new server just for this thesis.
• The original design for the user interface couldn’t cope with the fact that it’s impossible to design a page that’ll fit all mobile phones: there’s just too much variety in screensize and letter size. This was coped with by changing the design to allow the interface to run over the bottom of the screen.
• The most important problem arose when the brand-new Nokia 6210 Wap enabled mobile phone, costing 80 pounds and bought especially for this thesis, broke down. This resulted in much rushing to various repair centers in Oxford & London and eventually the use of a courier to get the phone picked up and fixed. In the meantime it had to be demonstrated to the supervisor that the server-side programs were working. How to get past this problem? This is where the seventh design principle of separation of concerns came in really handy. Because separate programs had been written to handle "translation", "calculation" & "page creation", it was relatively simple to output a HTML reply that could be viewed on the Web. This meant that it could be demonstrated that the server-side programs were working even though the phone was broken.