I added a fixed ebuild for sys-process/minit-0.10 to my overlay today, fixing various issues: 1. With the portage-ebuild, the build fails because being unable to find libowfat includes 2. Moved creation of /etc/minit/{in,out} FIFOs into pkg_setup, since paludis refuses to install FIFOs (good idea, IMHO) 3. Added some magic to get minit running on SPARC... On my E450 minit segfaults when built against dietlibc (wasn't able to examine if that's an issue concerning minit or dietlibc, guess latter). Added a patch fixing some build problems with official minit-0.10 tarball when building without dietlibc but glibc. Reproducible: Always
Please attach the ebuild to the bug.
Oh, maintainer-needed@? I could do it if no-one did mind.
Created attachment 191667 [details] fixed ebuild (build fixes, SPARC support) Sorry, read comment too late.
Created attachment 191669 [details, diff] diff used in ebuild Oh, somehow I always manage to get confused with bugzilla-usage. Here's the patch that's used in the ebuild. Ebuild on it's own doesn't fix all. Patch is mostly for fixing build against glibc but dietlibc. Most, when not all of this is fixed in current CVS revision. I already emailed Felix, asking when he's going to bring out a new release, but didn't get an answer so far. Sorry for the noise.
Hey, folks. I'm very disappointed, that your fix for this bug (and 248433) is hard-masking and scheduling the package for removal. The attached ebuild fixes 248433 as well. And concerning the missing maintainer... I already offered my services and I'm really puzzled that nobody here said anything about this offer - you could at least have said what's your point to decide I'm not the right one. _I_ don't see any problems with the package, besides you seeming to ignore my bug report. Just my 2¢. BTW: URL
I understand you are against the removal of this package, I'll give a look to the ebuild after watching bugs 262663 and 272507 (dev-libs/libowfat). If you are able to find any upstream webpage (or activity) that would be helpful to fix things faster.
I looked for activity in Fefe's CVS a while ago (when reporting this bug, AFAIR) and IIRC at least the issues reported in this bug are fixed in CVS. BTW: The build-issues aren't _really_ a problem of this package, but of libowfat's ebuild, since the header files aren't placed, where libowfat and depending packages think they are, but a sub-directory (good idea, indeed). You just have to remember that when writing ebuilds for packages that want to use libowfat.
Commited in portage, thanks for the patch. Feel free to open a keyword bug once you are able to avoid segfaults , or visit #gentoo-sparc at irc.freenode.org for some advice. Unmasking since I'll be a proxy maintainer of this package.
