Another set of autoconf macros for the purpose of building against boost.
Created attachment 260602 [details] boost-m4-0.1.ebuild
Created attachment 260604 [details, diff] files/boost-m4-0.1-multilib.patch Accepted upstream ( https://github.com/tsuna/boost.m4/tree/6e6036d31e00bd7f8e8170c78a3e6cde7597c287 ). Fixes finding 32-bit boost libs on multilib-nosymlink systems (i.e., on only my computer).
Committed into sunrise, will be available vie layman -a sunrise at the next review.
Created attachment 271973 [details] boost-m4-0.2.ebuild Updated ebuild for new version (v0.2). As mentioned the patch was accepted upstream, so this ebuild drops it. diff -u boost-m4-0.1.ebuild boost-m4-0.2.ebuild --- boost-m4-0.1.ebuild 2011-01-26 13:44:01.449055319 -0700 +++ boost-m4-0.2.ebuild 2011-05-03 06:19:11.654681817 -0600 @@ -24,10 +24,6 @@ mv * ${P} || die } -src_prepare() { - epatch "${FILESDIR}"/${P}-multilib.patch -} - # boost.m4 has a buildsystem which doesn't build nor install # anything. It only would add a buildtime dependency that boost be # installed. Thus, noop it:
Thanks for the notice, Erik. I updated the sunrise package and the new ebuild will show up at http://overlays.gentoo.org/proj/sunrise/browser/reviewed/sys-devel/boost-m4 whenever sunrise is reviewed again (within a week probably).
In gentoo-x86 now (thanks to mgorny's prompting ;-)): *boost-m4-0.2 (01 Feb 2012) 01 Feb 2012; Nathan Phillip Brink <binki@gentoo.org> +boost-m4-0.2.ebuild, +metadata.xml: Add boost-m4-0.2 for bug #352618, an alternate set of boost-detecting m4 macros.