…which is why there is so much badly-documented code out there, I suppose. As well as explaining why I seem to be spending time writing demos that no-one will probably ever use, rather than writing book chapters that no-one will ever read. Ah, what it is to have a meaningful life.
In my new job I’ll still be teaching, but I’ll no longer be teaching communications engineering (apart from one module on Internet Protocols), and I’ll be helping out in labs and workshops rather than doing conventional lectures. That changes the nature of the programs I’ll be writing; they’ll have to be intuitive for a student to use, and helpful for working things out in a lab, rather than just demonstrating ideas and concepts in lectures. So there will be more like the Lissajous demo, and less like the refraction demo. The next new one I’ve been thinking about will take a while though: I’ve been planning a digital circuit simulator where you can wire up gates and flip-flops and see what happens; plotting output traces and so on. This could turn into a major bit of software that’s useful in both lectures and labs.
However, I’ve still got all these old comms-related demos that I’d like to get updated and released before I get too involved in my new job. There’s a new one that I’ve just put on the web-site now – called Baseband Transmit. In the old VB6 days there was a corresponding receiver program that could calculate bit error rates, but I doubt I’ll get round to that again for Silverlight; it was always a bit too complex to use in lectures, and the only thing I used it for was demonstrating the performance of matched filters; I might write a simpler one just to do that. This one took a while, but then it’s the first one to plot graphs, and do any signal processing. Having got this going, the channel demo and Doppler demo should be much faster to convert over – a week or so each, I reckon. If I can stop myself working on the digital one, that is…