Hi Christian, Christian Kastner, on 2023-08-28: > On 2023-08-27 22:38, Étienne Mollier wrote: >> I did an upload of rocm-smi-lib to unstable today. If I have >> well understood the principle, the (lack of) test result should >> be reported by your infrastructure[1] in not too long. > > Correct. > > That didn't happen though, because there was another bug in the > scheduler. I forgot that even in a single distribution such as > 'unstable', multiple versions of a source package can exist. The > scheduler was assuming only one version, and returned after > version-comparing the first one. > > The fix was easy (check all versions), I already pushed it and the > package got picked up. Okay, now I see the CI run[1]. [1]: https://ci.rocm.debian.net/packages/r/rocm-smi-lib/unstable/amd64+gfx1030/ > [Incidentally, I don't know why rocm-smi-lib has multiple source > versions in unstable; it's not part of a transition AFAICS] Just conjecturing, but could it be a side effect of it being a key package? Cheers, :) -- .''`. Étienne Mollier <emollier@debian.org> : :' : gpg: 8f91 b227 c7d6 f2b1 948c 8236 793c f67e 8f0d 11da `. `' sent from /dev/pts/2, please excuse my verbosity `- on air: Frost* - The Boy Who Stood Still
Attachment:
signature.asc
Description: PGP signature