What are all these hypothetical humans doing with CSV files? I've worked with them a lot, and the only time I read one manually is to find out why it isn't parsing properly. I'm not writing them by hand. I'm not extracting data from them by hand. What are other people doing?
Sometimes you need to look up a single code in a 50 MB file. And sometimes you need a quick check to see if one line or a million lines changes.
It's "exception not the rule" type stuff... but it sure comes in handy to be able to check this stuff quick with basic text tools than have to run it through some binary parser. Same as JSON. But unlike protobufs for example.
Interesting. I would reach for `grep` and `diff` in those cases, so it's not much of a leap to imagine similarly using a tool for a binary format. After all, I only need the textual representation of the thing I'm searching for, not the entire dataset.
