Date: Thu, 18 Aug 2005 08:26:38 -0700 From: Brooks Davis <brooks@one-eyed-alien.net> To: Alexander Leidinger <Alexander@Leidinger.net> Cc: Brooks Davis <brooks@one-eyed-alien.net>, ports@freebsd.org Subject: Re: prebuild sanity checks Message-ID: <20050818152638.GA10363@odin.ac.hmc.edu> In-Reply-To: <20050818113949.kciryykkso00ko8k@netchild.homeip.net> References: <20050817195839.GA22027@odin.ac.hmc.edu> <20050818113949.kciryykkso00ko8k@netchild.homeip.net>
next in thread | previous in thread | raw e-mail | index | archive | help
--W/nzBZO5zC0uMSeA Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Content-Transfer-Encoding: quoted-printable On Thu, Aug 18, 2005 at 11:39:49AM +0200, Alexander Leidinger wrote: > Brooks Davis <brooks@one-eyed-alien.net> wrote: >=20 > >This started me wondering if we shouldn't have a few sanity checks in > >the build process so we refuse to build if the environment is missing > >some really critical things. Obviously, we can't test everything since > >it would take too long even if we had appropriate tests, but a few > >checks might save some hair pulling. For this particular case, I can > >think of two major ways to do it. First, we could just require that > >/dev/null exist to do anything. That's probably a bit intrusive though. >=20 > I don't think so. A lot of software depends upon it, e.g. configure scrip= ts > and even parts of our ports collection. Some parts may not produce unexpe= ted > results if it isn't available, even when it it used, but not having=20 > /dev/null > is a hack in this case and we shouldn't approve this misuse of the softwa= re. > So I think this isn't intrusive at all. Should I look at working up a patch? > >Another option might be a new variable (or variables) that ports that > >tend to break spectacularly and unobviously can set like: > > > >BUILD_DEVS=3D null zero > > > >Does this seem like a reasonable thing to do? >=20 > If you realy only talk about /dev/*: I object to make it a part of=20 > individual > makefiles. Either we depend on the common devs globally, or not at all. > Everything else is a maintainance nightmare (99% of the maintainers do/wi= ll > not test in such a restricted environment). OK, I'll forget about this idea. -- Brooks --=20 Any statement of the form "X is the one, true Y" is FALSE. PGP fingerprint 655D 519C 26A7 82E7 2529 9BF0 5D8E 8BE9 F238 1AD4 --W/nzBZO5zC0uMSeA Content-Type: application/pgp-signature Content-Disposition: inline -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.1 (GNU/Linux) iD8DBQFDBKiuXY6L6fI4GtQRAjQLAJ4gcRZ4fXWJv72thr7l8m+ztbA0JQCeNxak 3HYuvAiYPckSCcwbYShwQy4= =Lf6m -----END PGP SIGNATURE----- --W/nzBZO5zC0uMSeA--
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20050818152638.GA10363>