[Debichem-devel] [Debian-science-sagemath] jmol: diff for NMU version 12.2.32+dfsg2-1.1

[Ximin Luo]

Ximin Luo:
> Ximin Luo:
>> Andreas Tille:
>>> 4. Update Jmol in Debian, and update biojava4 and drop outdated jalview
>>>    (by at least starting the new dependencies and trying to upgrade
>>>    jalview) and have the chance of SageMath entering Debian stable.
>>>  
>>> [..]
>>
>> I agree and I'll be happy to move forward on this. Reminding everyone of the release schedule: https://lists.debian.org/debian-devel-announce/2016/11/msg00002.html
>>
>> - The deadline for packages-in-testing to be updated by a newer unstable version is ~Jan 25
>> - The deadline for new packages to enter testing, is ~Dec 25.
>>   - This includes packages that are in unstable, that were/will removed from testing for QA reasons. (e.g. Jmol 12)
>>   - This also includes time passing through the NEW queue, for new packages.
>>
>> In summary:
>>
>> - Jmol 14 must be uploaded to unstable by ~Dec 25
>> - biojava4, jalview must be uploaded to unstable by ~Jan 25
>> - jalview's extra dependencies must be uploaded to unstable (NEW) by ~Dec 20 or ideally earlier, to give FTP masters a few days to process it.
>>
> 
> Actually, we will have to upload Jmol 14 earlier, in order to "activate" the Jan 25 deadline for biojava4. This is because biojava4 is also QA-scheduled to be removed on Jan 4 (because of buggy-deps Jmol https://udd.debian.org/cgi-bin/autoremovals.cgi)
> 
> So we'd have to upload it on Dec 22 or earlier, to give some time for biojava4's AUTORM flag to get cancelled.
> 

Hi all,

I plan to upload Jmol 14 to unstable on Tuesday 20th December, sometime in the afternoon UTC.

Reasoning: We just uploaded SageMath to the NEW queue, aiming for experimental. Running some piuparts tests shows that it fails for the experimental suite. I didn't figure out the exact reason, but it is to do with apt-get not being able to figure out the correct combination of dependencies from experimental and unstable. I reproduced this in an empty schroot - I was unable to get "apt-get -fy" to install sagemath automatically from experimental, even with "--no-install-recommends". Also, "aptitude -R" failed to do this - it gives "remove sagemath" as the 1st preferred solution, and "install [the correct things]" as the 2nd option.

If this persists for the unstable suite, it will prevent SageMath from migrating to testing and stable. Therefore, we'd like some time to test SageMath-in-unstable before the Dec 25 deadline. It would make our lives much easier, to do these tests without having to reason about apt-get's special treatment of the experimental suite.

Consequences: I don't believe there are any negative consequences for the other packages. jalview, biojava3-live, biojava4-live are *already* scheduled to be removed on Jan 9. In fact uploading Jmol would reset this date to be further in the future, since it would close the RC bug that is causing this schedule. (Someone might then file FTBFS directly against these other packages, since they don't build against this new Jmol, but the counter would also be reset in that case.) There is a slight chance that I'm wrong about some/all of this, since I'm not *exactly* familiar with the release process, but this is based on what I understand from reading their documents and doing packaging for a few years.

If anyone doesn't want me to do this, please let me know before that date.

X

-- 
GPG: ed25519/56034877E1F87C35
GPG: rsa4096/1318EFAC5FBBDBCE
https://github.com/infinity0/pubkeys.git