On 2022-12-03 23:10:21 +0100, Andreas Tille wrote: > Am Sat, Dec 03, 2022 at 10:53:58PM +0100 schrieb Sebastian Ramacher: > > And there are many more. r-bioc-biocparallel is another one. This smells > > like many insufficient dependencies. > > Sorry, this all seems to be the same britney failure Paul was reporting. > The tests are all against the versions in testing which makes no sense > at all. These tests are not fixable since the old versions of BioC 3.15 > are bound to fail with packages from 3.16.
No, it's not. If you check the logs from r-bioc-htsfilter for example (https://ci.debian.net/data/autopkgtest/testing/amd64/r/r-bioc-htsfilter/28944996/log.gz), you see that some packages have actual regressions: Removing autopkgtest-satdep (0) ... autopkgtest [19:15:37]: test run-unit-test: [----------------------- cp: cannot stat '/usr/share/doc/r-bioc-htsfilter/tests/*': No such file or directory autopkgtest [19:15:37]: test run-unit-test: -----------------------] autopkgtest [19:15:37]: test run-unit-test: - - - - - - - - - - results - - - - - - - - - - run-unit-test FAIL non-zero exit status 1 autopkgtest [19:15:37]: test pkg-r-autopkgtest: preparing testbed > If this might help and is easily to realise removing all r-bioc-* > packages from testing could enable the migration of the packages in > unstable. I've done that now. Cheers -- Sebastian Ramacher
