On Mon, 2010-09-20 at 00:39 -0400, Trevor Woerner wrote: > The build wiki page on the fd.o site talks about building from > tarballs but suggests using the "build-from-tarballs.sh" script > instead of the "build.sh" script. Is the checkfortars() function code > that is being used in some way, or is it a left-over from something > else? >
I know that, generally speaking, people build from tarballs, but I am not aware if they do so from build.sh. The build-from-tarballs script hasn't seen its module list updated in 2 years. As hinted recently, this is duplication of modules list. When building from a tarball, I suppose, but I am not sure, that it is mostly for obtaining a previous version, or the latest published version as opposed to git master. There is no need to recreate the configuration, just running ./configure and make install. Note that autogen.sh is not part of the GNU packaging architecture and is not shipped in tarballs. If it helps, you can ignore that file and invoke "autoreconf -vfi" followed by ./configure --enable-maintainer-mode. You can find some background info on module configuration you can expect to be there as a result of project policy: http://wiki.x.org/wiki/NewModuleGuidelines To find out if the directory contains a fresh clone, you can check for the presence of some files generated by autoreconf. You can rely on Makefile.am to be present in all xorg modules root dir. It will help to inventory the files created as a result of run autoreconf vs ./configure. Keep in mind that the tarball generally gets created on one platform (say GNU/Linux) and gets configured and built on another (say Solaris). The developer scenario from git blurs the distinction between these steps. Gaetan
signature.asc
Description: This is a digitally signed message part
_______________________________________________ xorg-devel@lists.x.org: X.Org development Archives: http://lists.x.org/archives/xorg-devel Info: http://lists.x.org/mailman/listinfo/xorg-devel