Re: RFS: aegis (updated package, NMU)
CC'ing Christian, so that he is in loop.
>> - the orig.tar.gz you've used is content-identical but not
>> bit-identical. This means that the archive will reject the upload
>> because the checksums do not match.
> I'll regenerate the .changes using the proper archive.
> The problem may be related to the procedure used to produce the upstream
> tarball, it is created with file ownership different from root.root and
> git-buildpackage /probably/ is unable to reproduce it exactly.
git-buildpackage is able to reproduce exactly iff you use pristine-tar,
which is optional. In any case, I find the pristine-tar integration
pretty easy to use.
>> - Did you manage to talk about this NMU with Christian Meder? Does he
>> acknowledge this upload? In case you didn't manage to contact him, I
>> think you should adopt the package and use maintainer version
>> numbering. Espc. after reading #470162
> I don't know if he acknoledge the upload, but I've CCed him in my first
> RFS without receiving any response. I've also announced my intention to
> NMU on the BTS a week before the upload, without reactions.
> If I'm not wrong, you are suggesting I hijack the package (since it's
> not orphaned).
> Is it possible to do sponsored hijack?
You'd need to find a sponsor that signs the hijack with his key. Given
at the history of the aegis package, I think I'd be willing to do so.
Christian seems to be very busy since quite some time. While he did
answer to your most recent posts about aegis on firstname.lastname@example.org., as
far as I understand you, he has not gotten back to you.
If you follow up on the bug reports to obey the NMU protocol, I'd be
willing to upload your NMU at once. For an hijack, I'd upload to
10/day-delayed, so that Christian has some time to object to this upload
either by doing another upload in the mean time or by asking me to abort
If other debian developers disagree that this is a good way to go,
please tell me.
Reinhard Tartler, KeyID 945348A4