On 2013-09-09 15:44, Jérôme Vouillon wrote:
It seems Britney will happily remove from testing any binary packagetaken over in sid by another source package when its initial source package is removed or updated, instead of waiting for the new version of the package to beready. I'm wondering whether this is an expected behavior of Britney.
So far as I can see, you're misdiagnosing the cause(s) of the issues you're seeing. It has nothing to do with whether the binary packages have been "taken over", simply whether they exist in the version of the package migrating.
For instance, package proftpd-mod-geoip was removed from testingtogether with its source package on July 1, while the new version of the package build fromthe source package proftpd-dfsg is still stuck in sid.
This was because the source package was removed from unstable on the same day. When a source package is removed from unstable, britney will automatically try and remove it from testing as well; if there are no reverse-dependencies keeping it in testing then that will succeed.
Likewise, libcolord-gtk-dev and libcolord-gtk1 were removing from testing when the source package colord was updated on July 6;
The source package was "updated" to remove those binary packages. If the old packages have no dependencies in testing and are no longer built from any package in testing, then keeping them in testing would be wrong (as the source used to build them would no longer be in testing).
a new version of thesepackages were put back in testing the next day together with their new sourcepackage colord-gtk.Finally, ulogd and other packages from the source package ulogd were removed from testing on July 7, while transitional dummy packages of the same names remained in sid until July 15 together with the new source package ulogd2.
Again, the packages were removed from unstable on the date quoted, and had no reverse dependencies in testing, so there was nothing "holding" them in testing.
Regards, Adam