On Fri, Jan 14, 2011 at 05:10:23PM +0100, Guido Günther wrote: > On Fri, Jan 14, 2011 at 04:37:00PM +0100, Andreas Rottmann wrote: > > Package: git-buildpackage > > Version: 0.5.11 > > Severity: normal > > > > When git-buildpackage looks for a matching pristine-tar commit, it > > will use the earliest commit that matches. While normally, there > > should be only a single matching commit, there might be multiple when, > > for example, the initially checked-in original tarball was bogus. > > This happened to me when sponsoring failmalloc[0], where the initial > > pristine tarball was an accidentially repacked one. > > > > [0] git://git.debian.org/collab-maint/failmalloc.git > > > > Here is some output that should help to illustrate the issue: > > > > % git-buildpackage --git-verbose --git-export-dir=$PWD/../build-area -uc > > -us -S > > > > gbp:debug: ['git', 'branch', '--no-color'] > > gbp:debug: debuild -d clean [] [] > > dh clean --with autotools-dev > > dh_testdir > > dh_auto_clean > > dh_autotools-dev_restoreconfig > > dh_clean > > gbp:debug: ['git', 'status'] > > gbp:debug: ['git', 'branch', '--no-color'] > > gbp:debug: ['git', 'log', '--pretty=format:%H', '--grep=pristine-tar .* > > failmalloc_1.0\\.orig.tar\\.', 'pristine-tar', '--'] > > gbp:debug: Found pristine-tar commit at > > '363b271298c651f4deda9ce933833af5d76eacaf' > > [...] > > This commit is _only_ used to determine the compression type and looking > at the repo that's tar.gz for both commits. What issues are you seeing?
That said we should of course use the latest commit when guessing the compression type which is fixed in git. Cheers, -- Guido > > This looks more like pristine-tar picking up the wrong commit. > Cheers, > -- Guido > -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
