On Sun, Mar 29, 2026 at 03:53:08PM +0200, Matthias Klose wrote: > > During a rebuild of all packages in unstable, this package failed to build. > > the build just succeeded on the buildds
By pure chance. > plus it succeeds locally By pure chance. > So no, it builds. By pure chance. Packages that build ok by pure chance may not be said to build from source. > If you have a buildd configuration that differs from the ones used on the > Debian buildds, you should mention it in the bug report, You can reproduce this easily by doing this before the build: export DEB_BUILD_OPTIONS=parallel=1 which clearly shows the bug in the Makefile. However, that does *not* mean that it will *only* fail that way. The precise cases where it fails are still unknown, but given that there is a patch fixing the issue on the table, we don't really need to know. > and not exaggerating the severity of the report. Are you claiming this is not a violation of Policy 4.2? Please quote chapter and verse in Policy where it says this is wishlist, as I don't think I am the one exaggerating severities. It could be argued that this is also a violation of 4.9.1 regarding the implementation of parallel=n in DEB_BUILD_OPTIONS. Quote: If the package build system supports parallel builds, it must respect the parallel=n flag If the package build system does not support parallel builds, this string must be ignored. If the package build system only supports a lower level of concurrency than n, the package should be built using as many parallel processes as the package build system supports. Note that there are essentially two options here, namely "supporting parallel builds" and "not supporting parallel builds". The case where *only* parallel builds (with N>=2) are supported does not even exist, as it's taken for granted that packages will always build with make -j1. As far as I know, that's also the policy followed by the GNU project. Thanks.
