Date: Fri, 28 Aug 1998 18:27:14 +0000 From: Mike Smith <mike@smith.net.au> To: Matthew Dillon <dillon@backplane.com> Cc: Poul-Henning Kamp <phk@critter.freebsd.dk>, John Polstra <jdp@polstra.com>, committers@FreeBSD.org, jkh@time.cdrom.com Subject: Re: make.conf Message-ID: <199808281827.SAA02369@dingo.cdrom.com> In-Reply-To: Your message of "Fri, 28 Aug 1998 17:58:47 MST." <199808290058.RAA20957@apollo.backplane.com>
next in thread | previous in thread | raw e-mail | index | archive | help
> had no proble removing the -c portion of the ping commit. I have no > problem discussing the make.conf localization issue that I would like > to add to-current... an issue, I might add that, that isn't something > I thought up in 10 seconds and decided to commit. We've been working > with make.conf for well over 2 years and the relatively innocuous feature > request and commit came out of those 2 years in working with it. Just out of curiosity, given that /etc/make.conf is an entirely "local" (ie. site-specific) file, if you want to localise such that all your systems read an override file, what's to stop you adding the include to your 'site-standard' /etc/make.conf, rather than making it part of system policy at all? Adding make.conf.local to sys.mk only gives you one level of nesting, in a hierarchy that is entirely local to begin with. Embedding your own nesting structure in the hierarchy would seem to make more sense, but I'm not sure that it's useful in the general case. When it comes to this sort of decision, it's always worth remembering that not everybody else is addressing the same set of problems that you are. Your justifications may well be perfectly reasonable in your context, but they're not going to be obvious to people that are meeting your proposal for the first time. With something like this three-liner, the concern that comes to most people is not "is this a good idea", but "will this idea rot, undocumented, like so many other minimal hacks"? Is there a bigger picture? Is it perhaps amenable to a different solution with wider applicability? Is a site-specific customisation more effective? These are all questions that I don't think have been resolved, and this may be why consensus hasn't been reached. -- \\ Sometimes you're ahead, \\ Mike Smith \\ sometimes you're behind. \\ mike@smith.net.au \\ The race is long, and in the \\ msmith@freebsd.org \\ end it's only with yourself. \\ msmith@cdrom.com
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199808281827.SAA02369>