Hi,
On 19 May 2022 1:10:23 pm IST, Paul Gevers <elb...@debian.org> wrote:
>Package: release.debian.org
>User: release.debian....@packages.debian.org
>Usertag: britney
>
>Hi,
>
>On 11-05-2022 12:34, Graham Inggs wrote:
>> Yes, please go ahead.
>>
>> Tracker is at:
>> https://release.debian.org/transitions/html/r-api-bioc-3.15.html
>
>There's three package still waiting for an upload, r-bioc-progeny,
>r-bioc-monocle and r-bioc-scater. When is the ETA for those?
They are blocked by packages in NEW so whenever FTP masters accept these.
I think we can workaround these anyway by editing the DESCRIPTION file -- maybe
I'll do so and upload either late in the night today or early morning Tom.
>Apart from these missing uploads, there's a bunch of autopkgtest regressions
>blocking the migration of the set [1]. I have a hunch that quite some
>regressions are caused by missing *versioned* (test) dependencies that cause
>incompatible sets to be tested.
Many of these are also due to versioned depends I guess.
> Thinking more about it that could very well be a bug in our migration
> software failing to properly handle the Provides which glues this all
> together and which is uniquely provided by r-bioc-biocgenerics. Remember that
> our migration software tries to determine the minimal (source based) [...]
I know, if that could be improved, it'd be awesome otherwise it's a pain to
deal with. Even for an earlier r-base transition, I had to do bogus uploads to
make Britney happy.
It does not make much sense to me to add a breaks for things and clutter
d/control. There's no real usecase to be honest since there's already a version
dep on r/r-api-bioc respectively.
> if the manual pinning helps, we'll need to see how to improve it.
Thanks Paul for your work!
>[1]
https://people.debian.org/~elbrus/ci/regressions.html
>[2]
>https://ci.debian.net/data/autopkgtest/testing/amd64/r/r-bioc-megadepth/21776746/log.gz
--
Best,
Nilesh
