Ask HN: When startup time for CLI tools starts bothering you?

I am developing a CLI tools for internal usage to substitute the old tooling that we have, and one of the things that I am trying to optimize is startup time.

Why? Because with the old tools, every time that one command took more than a second I would unconsciously avoid this command, even if it was useful (of course, there are exceptions, i.e.: commands that are rarely used).

However, I only have arbitrary values for benchmark based on startup times that irritates me. This is basically what I am using as a base to measure against:

- ~0.1s: good, feels instantaneous

- ~0.2s: ok, however the latency starts to be perceptible

- ~0.5s: slow but bearable

- ~1s: I actively avoid to use this command unless necessary

A very simple test that you can do in your shell:

     foo() { sleep 0.1 && echo "Hello" }

And afterwards, call `foo`. You can of course play with `sleep` values to simulate other startup times.

Remember that I am looking at startup time, so it is the time until the commands gives some feedback. The command itself can takes more time than +10s to run, for example, however if it gives feedback in ~0.1s it still feels good.