Date: Wed, 30 Jan 2002 16:12:38 -0600 From: "Mike Meyer" <mwm-dated-1012860759.4ab381@mired.org> To: Michael Smith <msmith@freebsd.org> Cc: stable@freebsd.org Subject: Re: *_enable="YES" behavior is bogus Message-ID: <15448.28630.762965.100647@guru.mired.org> In-Reply-To: <200201301000.g0UA0Yb01251@mass.dis.org> References: <15447.22597.666281.179771@guru.mired.org> <200201301000.g0UA0Yb01251@mass.dis.org>
next in thread | previous in thread | raw e-mail | index | archive | help
Michael Smith <msmith@freebsd.org> types: > > If you build a kernel without INET, the following switches don't > > result in any daemons being started if you set them to YES: > > > > syslog_enable > > > sendmail_enable > > lpd_enable > > I think you are confused. These will all start regardless of whether > INET is enabled. They all may have started, but they exited immediately and were not enabled. > > These are all things that are clearly useful on a system without an > > internet connection. None of them start if you disable INET in the > > kernel. Neither the defaults/rc.conf file, the rc.conf man page, nor > > the man pages for those daemons mentiont that INET is required in the > > kernel to run them. Nor do the LINT or GENERIC files mention that > > turning off INET will cause various daemons useful on a machine > > without an internet connection to quit working. > You seem surprised that programs which depend upon networking don't > function as expected when it is not available. You seem to have missed the point of the posting, which is that expecting the rc.conf variables to have the same results no matter how you compile the kernel is unreasonable. > > If anyone is actually willing to > > look at fixing these things, I'll PR them. And probably start a > > witchhunt for other rc.conf variables that can be broken by careful > > selection of options in the config file. > Why not fix them yourself? The issues you're observing have nothing to > do with rc.conf variables; they have to do with bugs in the > above-referenced programs, specifically, they depend upon networking. > > > So simply > > expanding the documentation and comments will solve the problem. > No, it will not. Making the programs robust (where reasonable) is > necessary. Actually, syslog and lpd both have options to disable opening tcp sockets. Enabling those options would solve the problem. It's not clear what sendmail should do under those conditions. <mike -- Mike Meyer <mwm@mired.org> http://www.mired.org/home/mwm/ Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-stable" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?15448.28630.762965.100647>