Date: Wed, 14 Dec 2005 14:48:56 +0100 From: Alexander Leidinger <netchild@FreeBSD.org> To: Ulrich Spoerlein <q@galgenberg.net> Cc: emulation@FreeBSD.org Subject: Re: [PATCH] Fix excessive RUN_DEPENDS on archivers/rpm Message-ID: <20051214144856.o2p5blzyascwckcc@netchild.homeip.net> In-Reply-To: <20051214104108.GA1184@galgenberg.net> References: <200512132014.jBDKE0QG095682@roadrunner.q.local> <20051214101631.x0ca6keuo0owow8w@netchild.homeip.net> <20051214104108.GA1184@galgenberg.net>
next in thread | previous in thread | raw e-mail | index | archive | help
Ulrich Spoerlein <q@galgenberg.net> wrote: > Alexander Leidinger wrote: >> I agree that the RUN_DEPENDS is not ok, but: >> >> >Index: linux-gtk2/Makefile >> >> This needs a working linux-glib2 at install time, since it creates a list of >> gtk2 modules with the <mumble>query-modules program. By removing the >> BUILD_DEPENDS of linux-glib2 it's not possible to generate the list on the >> build cluster. > > But RUN_DEPENDS get satisified _before_ the do-install target. At least > on my system glib2 will be installed, before the actual installation of > gtk2. > > Since these ports set NO_BUILD and their build-stage is empty, > BUILD_DEPENDS is aequivalent to RUN_DEPENDS. At least that's how I see > it. I've looked at the CVS history... I think I was a little bit confused. I thought I added some parts to build_depends, but I added to run_depends (or maybe it was an interim version which did it in do-build... it's nearly a year ago, I don't remember). You are right, it should be enough to have run_depends, no need for build_depends here. Thanks for looking at this. Bye, Alexander. -- http://www.Leidinger.net/ Alexander @ Leidinger.net: PGP ID = B0063FE7 http://www.FreeBSD.org/ netchild @ FreeBSD.org : PGP ID = 72077137 She been married so many times she got rice marks all over her face. -- Tom Waits
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20051214144856.o2p5blzyascwckcc>