On Thu, Mar 31, 2011 at 04:19:55PM -0700, Russ Allbery wrote: > Roberto C. Sánchez <robe...@connexer.com> writes: > > > I am not sure precisely what you mean. I created a fresh checkout on a > > different machine (the first build was on amd64, the second on i386) and > > build again using svn-buildpackage. The result of the second build was > > the same as the first: > > > N: Processing binary package tbb-examples (version 3.0+r147-2) ... > > W: tbb-examples: gz-file-not-gzip > > usr/share/doc/tbb-examples/examples/parallel_reduce/convex_hull/convex_hull_bench.cpp.gz > > > In both instances, I ran lintian from inside a sid chroot. > > Okay, hm. That does seem like some sort of separate bug. > > What I was asking about before was if you had tried unpacking the built > *.deb file inside the chroot and then run the file program inside the > chroot on the *.deb file to see if it returned the correct recognition of > a gzip file or if it returned something else. Lintian should just be > calling file. > > If the file program in the sid chroot is broken, then this is probably the > same as the existing file bug. If the file that Lintian is using (the one > in the chroot) says it's a gzip'd file, but Lintian thinks it isn't, then > there's some other Lintian bug. > OK. After unpacking in the chroot and running file, here is what I get:
(unstable)roberto@miami:/var/lib/chroot/pbuilder-sid/results/tmp$ file usr/share/doc/tbb-examples/examples/parallel_reduce/convex_hull/convex_hull_bench.cpp.gz usr/share/doc/tbb-examples/examples/parallel_reduce/convex_hull/convex_hull_bench.cpp.gz: Minix filesystem, V2, 30 char names, 52542 zones Running it outside of the chroot on that same system (running Squeeze) yields the same result. Running it on the system where I first encountered the bug (running Lenny still) yields the results I provided in the earlier mail. Regards, -Roberto -- Roberto C. Sánchez http://people.connexer.com/~roberto http://www.connexer.com
signature.asc
Description: Digital signature