See https://www.debian.org/Bugs/server-control#affects If you fail to reproduce this, please provide a build log and diff it with mine so that we can identify if something relevant changed in the meantime.
Using non-free-firmware archive area did help! (my build fails at the moment, due to some pulseaudio-module-zeroconf dependency issue, but that is probably just transient issue in the archive, not live-build issue).
Unfortunately when fakeroot fails it results in results in the wrong file permissions in the .deb, but in a production environment it is a matter of tracking down which build host the broken build came from.
Source: python-transitions Version: 0.8.1-1 Severity: serious Justification: FTBFS on amd64 Tags: bullseye sid ftbfs Usertags: ftbfs-20200620 ftbfs-bullseye Hi, During a rebuild of all packages in sid, your package failed to build on amd64.
I expected the `broadcom-sta-dkms` module to build successfully for kernel `6.12.12+bpo-amd64`, allowing the kernel and dependent packages (`linux-image-amd64`, `linux-headers-6.12.12+bpo-amd64`, etc.) to configure and complete the upgrade without errors.
Every failed build was retried once to eliminate random failures. Added blocking bug (s) of 964663: 964724 Request was from mjt <[email protected]> to control@bugs.debian.org. (Thu, 09 Jul 2020 15:30:02 GMT) (full text, mbox, link).
It's vital for "dpkg-buildpackage -S" -- many of us use sbuild for binary builds so build-dependencies are unlikely to be installed on the host. It's especially annoying if you do porting or qa as you handle many many packages.
And unfortunately the help page (https://lintian.debian.org/tags/missing-build-dependency-for-dh-addon) doesn't help much either. For me the solution was to remove `--with autoreconf` from d/rules.
* Implement the required build-arch and build-indep targets in debian/rules. (Closes: # 998978) * Make the build reproducible by adding "-n" to the gzip (1) invocation.
