Date: Wed, 21 Jun 2006 16:07:06 -0600 (MDT) From: "M. Warner Losh" <imp@bsdimp.com> To: jhb@freebsd.org Cc: trhodes@freebsd.org, src-committers@freebsd.org, cvs-all@freebsd.org, obrien@freebsd.org, cvs-src@freebsd.org Subject: Re: cvs commit: src Makefile README Message-ID: <20060621.160706.1606922361.imp@bsdimp.com> In-Reply-To: <20060621.123456.232930587.imp@bsdimp.com> References: <20060620193458.42040d97.trhodes@FreeBSD.org> <200606211306.36770.jhb@freebsd.org> <20060621.123456.232930587.imp@bsdimp.com>
next in thread | previous in thread | raw e-mail | index | archive | help
In message: <20060621.123456.232930587.imp@bsdimp.com>
"M. Warner Losh" <imp@bsdimp.com> writes:
: In message: <200606211306.36770.jhb@freebsd.org>
: John Baldwin <jhb@freebsd.org> writes:
: : The nature of the compromise over 'make world' was that rather than removing
: : the knob we'd leave it undocumented so that new users aren't tempted to use
: : it, but instead will revert to using buildworld/installworld as they ought to
: : if they get a failure running 'make world'. If we are going to document the
: : knob and thus make it user-supported, we might as well remove it and just
: : make 'make world' user-supported.
:
: The problem with making it 'user-supported' is that it really is a
: dangerous tool for the normal user. Sure, it mostly works, most of
: the time, for most situations. However, when it fails, it fails in an
: unsafe way. It fails in a way that can't be backed out from easily,
: meaning someone will have to boot distribution media to back out to a
: known good state. These events are rare, but totally #@$@!!@@#$ the
: user.
:
: And we're back to the compromise that we had before...
Unless we want to require I_KNOW_WORLD_CAN_HURT_ME instead of
HISTORICAL_MAKE_WORLD :-)
Warner
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20060621.160706.1606922361.imp>