HEADS UP: set_rcvar() removed from rc.subr

Conrad J. Sabatier conrads at cox.net
Sun Jan 15 09:05:00 UTC 2012

On Sun, 15 Jan 2012 00:40:36 -0800
Doug Barton <dougb at freebsd.org> wrote:

> On 01/15/2012 00:11, Conrad J. Sabatier wrote:
> > Chris, if you're working on fixing ports' rc files
> Thanks for taking a look at this. FYI, all of the rc.d scripts that
> are actually in the ports tree have already been fixed. The outliers
> at this point are scripts that are included in the distfiles. Some of
> those even have patches in the tree already, and they were updated in
> the same sweep.
> While we certainly try to enforce <filename> == PROVIDE == $name for
> scripts that are in the tree, it can't be guaranteed that this is true
> for scripts that are not, so if name=foo, rcvar should be foo_enable.
> hth,
> Doug

Well, that's good news.  I wasn't sure how far along the work on this
had come yet.

I'm amazed in retrospect at the amount of discussion that took place
over this change, especially seeing as how the rcvar is hardcoded
elsewhere in these scripts anyway (I mean, how else to do
": ${foo_enable=NO}" without introducing yet another level of
indirection?).  No wonder you were feeling a little ticked off earlier.

I thought your point about the time spent editing vs. the number of
times a script will actually be run was right on target.

Keep up the good work!  :-)


Conrad J. Sabatier
conrads at cox.net

More information about the freebsd-current mailing list