Bug#1029417: gammu: FTBFS in bookworm (undeclared build-dependency on tzdata)
El 30/1/23 a las 12:36, Boian Bonev escribió: Before uploading I did purge tzdata from my sbuild chroot and the build succeeded. Probably you did a full build, with both arch-dependent packages and arch-independent packages. In such case sbuild installs all build-dependencies, the ones in Build-Depends and the ones in Build-Depends-Indep. That's probably why it worked. If you already have a clean chroot in which tzdata is not present, the following two things should both work: sbuild --arch-all --no-arch-any -d sid gammu sbuild --arch-any --no-arch-all -d sid gammu Currently, only the first command works, because tzdata is also required in the second command but it's only in Build-Depends-Indep. When adding or removing build-dependencies (this is not really related to tzdata), it is always a good idea to try those commands before upload, because we have different autobuilders for arch-independent packages and for arch-dependent packages and those two commands are what autobuilders really do. Please provide some log showing the error. I believe that with those additional explanations it should not be necessary. The way sbuild --arch-any --no-arch-all fails on a really clean chroot is exactly what I included in the very first email in this bug report. About the other things you mention: My summary would be that this is a bug according to current policy in bookworm (which I believe we likely will not change just before the release), and the fact that the bug is not RC just means that the bug is not so grave that we would want to remove the package if it's not fixed. While I am always willing to fix low priority stuff, I'd prefer to wait and see some final resolution before proceeding. So: It is up to you. I don't plan to NMU or anything like that, but I won't deny that I will be very happy if we could release bookworm without FTBFS bugs of any kind (which includes this one), or as few as possible. Thanks.
Bug#1029417: gammu: FTBFS in bookworm (undeclared build-dependency on tzdata)
Hello, Before uploading I did purge tzdata from my sbuild chroot and the build succeeded. This makes me think that maybe there is something else involved, but since I did not see the failure log, that is a wild uneducated guess. Please provide some log showing the error. I am silently following the discussion on debian-devel, silently because my feeling is as a DM I do not have a voice in this. What I see are formally valid arguments on one end and practically valid ones on the other. The main problem is that for parties not involved in the decision both sides are somewhat right and as a worst case example, someone else may start filing bugs to remove the tzdata deps. That would be perfectly valid, because packages will build on buildds without the tzdata dep and removing it would minimize the dep list. As a result the whole thing will lead to pointless work and only contribute to global warming. My unimportant opinion is that while minimizing the build chroots and using finer grained deps is a very good thing, the change should be agreed upon, formalized in policy, and made clear to everyone involved. Then announced when it will be implemented on buildds, then implemented. There are already way too many suboptimal and fuzzy things in Debian's processes, adding more is not good for the project and respectively its users. While I am always willing to fix low priority stuff, I'd prefer to wait and see some final resolution before proceeding. With best regards, b. signature.asc Description: This is a digitally signed message part
Bug#1029417: gammu: FTBFS in bookworm (undeclared build-dependency on tzdata)
found 1029417 1.42.0-6 thanks Hi. I see that tzdata was added to Build-Depends-Indep. That's not enough. The package now still fails to build (in a clean chroot) when building the arch-dependent packages (dpkg-buildpackage -B). The attached patch should fix this. Thanks.diff --git a/debian/control b/debian/control index 5471d04..74686c0 100644 --- a/debian/control +++ b/debian/control @@ -18,12 +18,12 @@ Build-Depends: libusb-1.0-0-dev [linux-any], pkg-config, sqlite3, + tzdata, unixodbc-dev, Build-Depends-Indep: doxygen, python3-breathe (>= 4.1.0), python3-sphinx (>= 1.3), - tzdata, Standards-Version: 4.6.2 Rules-Requires-Root: no Vcs-Browser: https://salsa.debian.org/debian/gammu
Bug#1029417: gammu: FTBFS in bookworm (undeclared build-dependency on tzdata)
Package: src:gammu Version: 1.42.0-5 Severity: serious Tags: ftbfs patch Dear maintainer: During a rebuild of all packages in bookworm, your package failed to build: [...] debian/rules binary-arch dh binary-arch --with sphinxdoc --buildsystem=cmake dh_update_autotools_config -a -O--buildsystem=cmake dh_autoreconf -a -O--buildsystem=cmake debian/rules override_dh_auto_configure make[1]: Entering directory '/<>' dh_auto_configure -- \ -DINSTALL_LIBDATA_DIR=/usr/lib/x86_64-linux-gnu \ -DSYSTEMD_SERVICES_INSTALL_DIR=/usr/lib/systemd/system \ -DINSTALL_LSB_INIT=ON \ -DSYSTEMD_FOUND=ON \ -DWITH_SYSTEMD=ON cd obj-x86_64-linux-gnu && cmake -DCMAKE_INSTALL_PREFIX=/usr -DCMAKE_BUILD_TYPE=None -DCMAKE_INSTALL_SYSCONFDIR=/etc -DCMAKE_INSTALL_LOCALSTATEDIR=/var -DCMAKE_EXPORT_NO_PACKAGE_REGISTRY=ON -DCMAKE_FIND_USE_PACKAGE_REGISTRY=OFF -DCMAKE_FIND_PACKAGE_NO_PACKAGE_REGISTRY=ON -DFETCHCONTENT_FULLY_DISCONNECTED=ON -DCMAKE_INSTALL_RUNSTATEDIR=/run -DCMAKE_SKIP_INSTALL_ALL_DEPENDENCY=ON "-GUnix Makefiles" -DCMAKE_VERBOSE_MAKEFILE=ON -DCMAKE_INSTALL_LIBDIR=lib/x86_64-linux-gnu -DINSTALL_LIBDATA_DIR=/usr/lib/x86_64-linux-gnu -DSYSTEMD_SERVICES_INSTALL_DIR=/usr/lib/systemd/system -DINSTALL_LSB_INIT=ON -DSYSTEMD_FOUND=ON -DWITH_SYSTEMD=ON .. [...] test 468 Start 468: test_sql_time 468: Test command: /<>/obj-x86_64-linux-gnu/tests/atgen/test_sql_time 468: Working Directory: /<>/obj-x86_64-linux-gnu/tests/atgen 468: Test timeout computed to be: 1500 468: Test "tm.tm_hour == 13" failed! 468: test_fill_time_t_gmt 468: test_fill_time_t_dst 468/880 Test #468: test_sql_time ...***Failed 0.01 sec Test "tm.tm_hour == 13" failed! test_fill_time_t_gmt test_fill_time_t_dst [...] The following tests FAILED: 468 - test_sql_time (Failed) Errors while running CTest make[2]: *** [Makefile:94: test] Error 8 make[2]: Leaving directory '/<>/obj-x86_64-linux-gnu' dh_auto_test: error: cd obj-x86_64-linux-gnu && make -j1 test ARGS\+=--verbose ARGS\+=-j1 returned exit code 2 make[1]: *** [debian/rules:21: override_dh_auto_test] Error 25 make[1]: Leaving directory '/<>' make: *** [debian/rules:12: binary-arch] Error 2 dpkg-buildpackage: error: debian/rules binary-arch subprocess returned exit status 2 (The above is just how the build ends and not necessarily the most relevant part) Note: I'm using the "patch" tag because there is an obvious fix (indicated in the subject). About the archive rebuild: The build was made using virtual machines from Hetzner, with enough memory, enough disk, and either one or two CPUs, using a reduced chroot with only build-essential packages (plus debhelper). If you could not reproduce the bug please contact me privately, as I am willing to provide ssh access to a virtual machine where the bug is fully reproducible. If this is really a bug in one of the build-depends, please use reassign and affects, so that this is still visible in the BTS web page for this package. Thanks.
