Cy Schubert

I'm using the FreeBSD ksh93-devel port, updated a day prior to the rollback (8cf92b28). I have no problems with "b". You might want to try the latest ksh2020 branch.

Original idea came from FreeBSD PR/265963 (https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=265963). Certainly more elegant than what the user suggested.

> Per your comment, threads might indeed be better (also easier to track for the likes of systemd that struggles with dual-process upsmon). Also systems-dependent though. Or a select loop...

First, my apologies for the poor quality pull request. I should have spent more time on it (though this isn't and shouldn't be my top priority ATM). Secondly, I've extracted...

Not quite there yet. The process attached to the terminal must itself fork() before forking children to disconnect from the terminal (or console). I'll test here before submitting an update....

I've made seven additional commits during testing and have rolled the patch into a pending FreeBSD port commit to sysutils/nut (2.8.0) and sysutils/nut-devel (tracking your master branch). https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=265963 has been...

I'm not sure how to resolve doc/nut.dict.

In foreground mode the parent (main process) simply forks for each logger child and waits for their completion. When the parent receives a SIGINT or SIGTERM the parent terminates all...

To explain fork()ing a little more, If in background mode it forks first. The new parent will fork the children and wait for them. If in foreground mode the original...

Call me old fashioned but I refuse to use the github app on my phone to review code. Code doesn't look right there. I only use the app for 2FA...