On Sat, Jul 22, 2017 at 7:28 AM, Adrian Bunk <b...@debian.org> wrote: > Source: golang-github-spf13-cobra > Version: 0.0~git20170314.0.7be4bed-1 > Severity: serious > Tags: buster sid > > Some recent change in unstable makes golang-github-spf13-cobra FTBFS: > > https://tests.reproducible-builds.org/debian/history/golang-github-spf13-cobra.html > https://tests.reproducible-builds.org/debian/rb-pkg/unstable/amd64/golang-github-spf13-cobra.html > > ... > === RUN TestChildCommandFlags > --- FAIL: TestChildCommandFlags (0.00s) > cobra_test.go:575: Wrong error message displayed, > invalid argument "10E" for "-i, --intone" flag: > strconv.ParseInt: parsing "10E": invalid syntax > ... > === RUN TestFlagsBeforeCommand > --- FAIL: TestFlagsBeforeCommand (0.00s) > cobra_test.go:994: Wrong error message displayed, > invalid argument "10E" for "-i, --intone" flag: > strconv.ParseInt: parsing "10E": invalid syntax > ... > dh_auto_test: go test -v -p 1 github.com/spf13/cobra > github.com/spf13/cobra/cobra github.com/spf13/cobra/cobra/cmd > github.com/spf13/cobra/doc returned exit code 1 > debian/rules:4: recipe for target 'build' failed > make: *** [build] Error 1
Hi Michael, As you can see from the above, these test errors have to do with an old version (from March 2017) of cobra, and TestChildCommandFlags and TestFlagsBeforeCommand have nothing to do with the test fixtures, but rather with cobra not yet catching up with its dependencies (updated versions). So I am not entirely sure if your fix in 0.0~git20170724.0.90fc11b-2, namely: * Add 002-update-test-fixture-paths.patch and removed DH_GOLANG_INSTALL_EXTRA to make build reproducible. (Closes: #869304) matches this issue. Besides, the previous version 0.0~git20170724.0.90fc11b-1 builds on all active architectures (except MIPS family due to gccgo using different GOARCH values than gc does), for example, s390x: https://buildd.debian.org/status/logs.php?pkg=golang-github-spf13-cobra&arch=s390x Unfortunately, tests.reproducible-builds.org does not keep any past logs, only a history list, and the FTBFS error I see now is just a "2017 <> 2018" mismatch because "build2" runs on a machine with a date set to 2018. So, if you don't mind, I would like to investigate more carefully by reverting your change in 0.0~git20170724.0.90fc11b-2 to see the what the original FTBFS error with 0.0~git20170724.0.90fc11b-1 on "build2" test machine actually was. Please refrain from making changes to this package in the coming week or two, as I will be experimenting with various fixes to get it compile on MIPS architectures too. Thank you for your understanding. Cheers, Anthony