Package: release.debian.org Severity: normal User: release.debian....@packages.debian.org Usertags: binnmu
I realise you don't want 3000 binNMU requests for the libstdc++ ABI transitions; but this particular one breaks the amd64 experimental buildds, which maintainers might be relying on for their ...v5 transitions, so it seems worth speeding up. Please rebuild cwidget against libsigc++-2.0-0v5, then rebuild aptitude against that: nmu cwidget_0.5.17-3.1 . ALL . -m "rebuild against libsigc++-2.0-0v5" nmu aptitude_0.7-1 . ALL . -m "rebuild against libsigc++-2.0-0v5" dw aptitude_0.7-1 . ALL . -m "libcwidget3v5 (>= 0.5.17-3.1+b1)" This is currently affecting all the amd64 experimental buildds, for instance <https://buildd.debian.org/status/logs.php?pkg=dbus&ver=1.9.20-1&arch=amd64> (which was successful on all other architectures). However, if there's some subtle ABI issue, it seems safer to rebuild everywhere. I can reproduce the sbuild failure locally when using "--build-dep-resolver=aptitude", which I think is what the official experimental buildds do. When I rebuild cwidget in an unstable sbuild, rebuild aptitude with the new cwidget added via --extra-package, then build dbus in an experimental sbuild with the updated cwidget and aptitude, the dbus build succeeds. New cwidget + old aptitude is not sufficient (aptitude becomes uninstallable). Fixing this should hopefully mean people can use pbuilder again, too. Regards, S -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
