Date: Fri, 11 Sep 2015 18:11:12 +0100 From: RW <rwmaillists@googlemail.com> To: ports@freebsd.org Subject: Re: rc script problem - pidfile not being recognised Message-ID: <20150911181112.04651c9e@gumby.homeunix.com> In-Reply-To: <55f30648.b343320a.52a32.0690SMTPIN_ADDED_MISSING@mx.google.com> References: <20150911161850.402b3d5a@gumby.homeunix.com> <853047065.182.1441986957380.JavaMail.Kevin@Thoth> <20150911171930.59e67cc2@gumby.homeunix.com> <55f30648.b343320a.52a32.0690SMTPIN_ADDED_MISSING@mx.google.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Fri, 11 Sep 2015 09:46:31 -0700 (PDT) Roger Marquis wrote: > RW via freebsd-ports wrote: > > It needs both. It won't use just the pid file because the pid > > might have been reassigned to another process if the original daemon > > died without deleting its pid file. > > Why would this rc script *require* a command_interpreter variable to > use the pidfile variable? I'm curious because this violates KIS, the > principle of least surprise and few rc scripts seem to have this > variable defined. > > Using command_interpreter is good to be sure, for for the reason > listed, but rc scripts should not fail if it is undefined. IIRC the name is always checked. When an interpreted script is run using a shebang the command in the output of ps doesn't match the command used to start the script, so command_interpreter is needed to get a match.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20150911181112.04651c9e>