I frequently find myself needing to rerun commands with the -h flag, and I was just wondering why it is not the default behavior of programs to show sizes in human-readable terms?
I think it’s two reasons - first is that human readable formats are less specific, so three files that are “3gb” in size look the same, so it may be hiding information that it knows about
The second is that commands are very often used as parts of larger scripts, where human readability may not matter as much, or may not be applicable until the very last command when the output is actually shown.
Between these, and the fact that most commands try not to predict what your use case is, I think that having a -h option is going to be the norm going forward as well rather than having it as the default
I guess because most of the time the data is consumed by another script. I don’t know how OpenBSD behaves, but I would be pretty angry if I had to put such flags (like “-not-human”) everywhere in my code, docker scripts, and pipelines.
