Totally fair. I had meant that to be only concerning complicated assembly. Either some of the more obscure instructions or SIMD, in general.
I think my more general idea, that I was vaguely grasping towards, is that the style of programming that produces optimal assembly is already a different style than most people use. Certainly a different one than most people talk about using.
And fair on the CSV points. I definitely meant that to be in numeric heavy CSV files. Strings are a whole other beast. I put https://taeric.github.io/challenging-my-filesize-intuition.h... together a few years ago when I had this discussion at work the last time.
Oh, another one: the byte length of my strings is the same as the character length of my strings. That one is so nasty, especially in a UTF-8 setting where if you naively test with ASCII strings the tests will all be written with that assumption in mind.
I think my more general idea, that I was vaguely grasping towards, is that the style of programming that produces optimal assembly is already a different style than most people use. Certainly a different one than most people talk about using.
And fair on the CSV points. I definitely meant that to be in numeric heavy CSV files. Strings are a whole other beast. I put https://taeric.github.io/challenging-my-filesize-intuition.h... together a few years ago when I had this discussion at work the last time.