We spend a great deal of time in computer science worrying about computer languages, their constituent symbols and their relative worth.
In and amongst all this angst, we don’t seem to have learned very much. We have a few new languages each decade to get excited about, but we still produce software which even charitably can be awful for no other reason than the management process which produced it. Sometimes it’s not only awful but lethal - you only have to think as far as the Boeing 737 MAX debacle which brought the whole concept of safety-critical software into disrepute. We can devote serious intellectual time to a particular software feature only to find the whole thing has been undermined by a spreadsheet jockey.
In 2020, my car got stolen because some “coder” had decided that keyless was the way to go. Perhaps that should be clueless. In 2022, my neighbour’s car (a different manufacturer) was stolen the same way. It takes around a minute. I expect the perpetrators of these software masterpieces will be busy on self-driving cars by now. Oh joy. Last week, I volunteered to enter results for my local athletics club only to find that a new client-server system had been introduced to “facilitate” this. I spent the afternoon sifting through web server “5xx” messages before sliding unconscious under the table. Will we ever learn anything? Do I even care anymore?
Since we clearly can’t deal with systems of symbols that mean something, what can we learn about systems of symbols when they have no meaning?
Turns out to be rather a lot.