Re: get-gauche.sh shouldn't use make -j without options

Shiro Kawai <[email protected]> writes:

> Is nproc enough to prevent disruption?  (I mean, should it be less than
> nproc to keep a few cores for other processes?

nproc should be good enough. I also think my new audio setup using
pipewire is more sensitive to this, than it used to be (imho ideally
music wouldn't be interrupted no matter how high the system load is)

Regarding portability:

- on debian systems nproc is in the coreutils package which is described
  as:
  "
  This package contains the basic file, shell and text manipulation
  utilities which are expected to exist on every operating system.
  "

- looks like freebsd also has nproc (but is that installed by default?):
  https://man.freebsd.org/cgi/man.cgi?query=nproc&sektion=1&format=html

did some more search and found:

a) https://stackoverflow.com/questions/45181115/portable-way-to-find-the-number-of-processors-cpus-in-a-shell-script
suggesting "sysctl -n hw.ncpu" on BSDs

b) https://unix.stackexchange.com/questions/208568/how-to-determine-the-maximum-number-to-pass-to-make-j-option
interesting read ... looks like there are machines with not enough RAM
to run big compilations on all CPUs in parallel
=> in debian there is even a package to guess the value for make -j:
https://sources.debian.org/src/guess-concurrency/0%2B20250305/guess_concurrency.py

> I'm usually not bothered by this so I don't know.)

so many CPUs?