Re: RFS: macaulay 1.17.1

2021-01-18 Thread Torrance, Douglas

On 1/17/21 2:13 PM, Doug Torrance wrote:

On 1/17/21 8:57 AM, Anton Gladky wrote:

Hi, I will take care of it.


Thanks, Anton!

I noticed that Andreas uploaded it earlier today (thanks, Andreas!), but 
it was an old draft from the "master" branch that I had forgotten to 
delete from Salsa after switching to "debian/master" as the default 
branch a couple months ago.  I merged that into debian/master (for the 
d/changelog entry) and deleted the old branch to avoid further 
confusion.  So if you'd be able to upload 1.17.1 when you get a chance, 
that would be great!


Thanks again for sponsoring, Anton!

The autopkgtest tests are failing on the 32-bit architectures, and 
they're being listed as regressions for the purposes of migration to 
testing [1].  However, they've only previously succeeded in 
experimental, and according to [2], regressions are only supposed to 
block migration when they've previously succeeded in testing.  So I'm 
confused...


(The regression is due to some new tests that were added between the 
version in experimental and the version that was just uploaded to 
unstable.  We run out of memory during these tests on the 32-bit 
architectures.)


Does anyone know which is correct?  I'm working on a patch that I think 
should fix it, but I'm not sure if it will be necessary to upload a new 
version in order to migrate or not.


Thanks!
Doug

[1] https://qa.debian.org/excuses.php?package=macaulay2
[2] https://wiki.debian.org/ContinuousIntegration/DebianSetup


OpenPGP_0xD12B2BE26D3FF663.asc
Description: application/pgp-keys


OpenPGP_signature
Description: OpenPGP digital signature


Re: Sponsoring before freeze

2021-01-18 Thread Anton Gladky
Hi Thomas,

uploaded! Thanks for your contribution.

Please consider adding autopkgtest for the next upload. If you
prepare them before freeze, please let me know and I will upload it.

Regards

Anton

Am Do., 14. Jan. 2021 um 00:39 Uhr schrieb Thomas Schiex
:
>
> Dear Anton,
>
> I have tried to prepare toulbar2 with a new upstream version. If something is 
> missing I'm happy to inject extra effort in it.
>
> Thomas



Re: python-rioxarray

2021-01-18 Thread Anton Gladky
Hi Magnus,

there are only recommendation, how do I usually deal with
such kind of problems:

> * the upstream tar ball comes with the egg-info directory. Presumably
> I should remove that since it gets generated?

Yes, remove. Put the wildcard in d/copyright under Files-Excluded section
and do "uscan --download --repack".

>  * the upstream project helpfully credits parts of the code that they
> have adapted from other projects. However, they folded their changes
> into other files. How is that reflected in the copyrights file?

All copyrights-holders should definitely be mentioned in d/copytight.
Otherwise the package will be rejected. The best way to drop all embedded
copies of third-party code and replace them by packaged versions, if they are
available in archive and if it is technically possible.

>  * I do get a bunch of warnings about privacy issues in the sphinx
> generated docs. I attempted to use the same rules as xarray but that
> didn't quite work. Also there are references to other images. How are
> these dealt with?

I am using "sed" to drop those places [1], [2]. But it can sometimes break
images, links etc.

[1] https://salsa.debian.org/science-team/yade/-/blob/master/debian/rules#L66
[2] https://salsa.debian.org/science-team/sumo/-/blob/master/debian/rules#L29

Regards

Anton

Am Do., 14. Jan. 2021 um 19:39 Uhr schrieb Magnus HAGDORN
:
>
> Hi all,
> I have just uploaded the first of my python packages that is probably
> closest to being finished to mentors. I am still in the process of
> figuring out how debian packaging in general and python packaging in
> particular works. I did copy a lot of the debian build machinery from
> the xarray package which was very helpful.
>
> So, I have a few issues still:
>  * the upstream tar ball comes with the egg-info directory. Presumably
> I should remove that since it gets generated?
>
>  * the upstream project helpfully credits parts of the code that they
> have adapted from other projects. However, they folded their changes
> into other files. How is that reflected in the copyrights file?
>
>  * I do get a bunch of warnings about privacy issues in the sphinx
> generated docs. I attempted to use the same rules as xarray but that
> didn't quite work. Also there are references to other images. How are
> these dealt with?
>
>  * I need to sort out the tests
>
> I have added the repo to salsa, it can be found here:
> https://salsa.debian.org/science-team/python-rioxarray
>
> Cheers
> magnus
> The University of Edinburgh is a charitable body, registered in Scotland, 
> with registration number SC005336.