2 ms·
Anyone who brings up sysvinit and the "30x 50-line shell scripts" is clearly one whose perception is limited to the ad-hoc LSB-based sysvinit setups that once p
by vezzy-fnord 11y ago
Anyone who brings up sysvinit and the "30x 50-line shell scripts" is clearly one whose perception is limited to the ad-hoc LSB-based sysvinit setups that once permeated GNU/Linux, but which have always been merely one relatively inefficient approach. Even back when Debian first introduced their sysv-rc, we already had IBM SRC, daemontools, simpleinit and NetBSD rc.d which showed different solutions.
EDIT: Let me join in on the moaning about downvotes. Not that I understand why.
- nailer 11y agoSysVinit is systemd's predecessor, and thus the logical comparison. Upstart would also be a relevant, if anyone had ever adopted it properly, unfortunately it never got past being a SysVinit wrapper. I've used minimal shell based systems on BSD, Solaris SMF, and launchd, as well as Windows services, and maintain the successor to the now-unmaintained Unix Rosetta Stone. However a minimal shell-script based init is not something anyone on Linux is seriously considering, so I'm not going to debate it with you - that would be a waste of both our time. But please, feel free to tell me about my own experiences. Since you're new here, have you read https://news.ycombinator.com/newsguidelines.html https://news.ycombinator.com/newsguidelines.html yet?