Why SIGQUIT for kill?

[stevehartdata]

Thank you for your work on this package!

Is there a reason you use SIGQUIT for kill and killAndWait? The default behavior on Linux is to terminate the process and dump core on SIGQUIT. I think it might make more sense to use SIGTERM, which is the default for kill(1). Would you be inclined to accept a PR to switch to SIGTERM?

[msrdic]

I am also interested if there's a reason behind it. While trying out the library, both kill and killAndWait didn't shut down my process. Changing it to sigTERM worked great.

[Javran]

I want to add that there are situations that GHC installs handler for SIGQUIT, which prevents process from termination: https://gitlab.haskell.org/ghc/ghc/-/issues/17451

[Javran]

Urgh, I was revisiting this and realized SIGTERM might not guarantee bracket exit action to be executed: https://gitlab.haskell.org/ghc/ghc/-/issues/10169