I worked late tonight for no good reason. My deadlines are all self-imposed. I just got a little excited, missed the reasonably-timed buses, caught a late bus back. Mother Nature abetted my bad behavior, displaying a full moon over a rippling San Francisco bay, a nice view for an evening commute. Sometimes I get a little excited. Blame the full moon.
A while back, I read three Tufte books in the space of a couple of weeks. He writes about conveying data through graphics for E-Z analysis. I was bouncing up and down with excitement about charts. I wanted to draw a chart. But I didn't know what data I should try to present. Finally, I chose something gratuitous.
When designing a language like Morse code, how do you choose the encoding? It makes sense to use short symbols to transmit common letters, and long symbols to transmit the rare letters. How well did the designers of International Morse Code achieve this goal?
(The numbers behind this chart are arguably contrived. I say that dashes take three times as long to transmit as dots. But that's only true if you're keying these letters by hand. There are (were?) plenty of telegraph systems that used the same amount of time to transmit "dit" as they did "dah". But that would have been less fun to chart.)
Labels: book, decoration