Descriptive names considered harmful
Ten years ago I was able to understand mathematical texts so much better than I’m able to now. And it is neither AD, nor ADD, or tequila that I blame. I blame descriptive identifier names in program code. I have become spoiled by them. I expect to see long and descriptive identifiers in equations rather than single letters with overload resolution based on typography. And since I’m also spoiled by intelligent code editors, I expect to be able to trace a symbol’s definition without having to go five pages back and re-reading them.
For instance, I haven’t been able to understand what the heck the Einstein field equation really means, no matter how many times and how many texts I read. Let me tell you that by convention, a boldface R is the Riemann curvature. But sometimes it’s written as Rabcd. And Rab is the Ricci curvature. But when you see just an R (not bold and without indices) – it is the scalar curvature. Needless to say that all those Rs are closely related, and if you confuse some of them just for a second – you fail.
So in order to restore justice I want a programming language that differentiates names by typography (bold, italic, script, double struck, lower and upper indexes, Greek and Gothic alphabets, some Hebrew too) and has context-based overload resolution. And it should be named ℂω or something.