[Debian-med-packaging] Bug#845737: drop the problematic testcase?

[Andreas Tille]

On Tue, Dec 13, 2016 at 02:20:39PM +0000, Ghislain Vaillant wrote:
> I would just drop the `test_symlink_time_handling` testcase and check
> with upstream what might be going on here.
> 
> Apart from that, it looks like the internet access errors are fixed.

I enabled the test passing instead of droping it.  I somehow have the
feeling that the "file exists" in the current build log

...
4 of 4 steps (100%) done
..:
total 4
-rw-r--r-- 1 pbuilder pbuilder 2496 Dec 13 15:25 Snakefile
-rw-r--r-- 1 pbuilder pbuilder    0 Dec 13 14:25 input_file
lrwxrwxrwx 1 pbuilder pbuilder   10 Dec 13 11:25 input_link -> input_file
lrwxrwxrwx 1 pbuilder pbuilder   10 Dec 13 13:25 output_link -> input_link
Provided cores: 3
Rules claiming more threads will be scaled down.
Job counts:
	count	jobs
	1	main
	1	make_output
	2
rule make_output:
    input: input_link
    output: output_link

ln: failed to create symbolic link 'input_link': File exists
^[[31mCalledProcessError in line 37 of /build/snakemake-3.9.0+dfsg/.pybuild/pythonX.Y_3.5/build/tests/test_symlink_time_handling/Snakefile:
Command 'ln -s input_file input_link' returned non-zero exit status 1
  File "/build/snakemake-3.9.0+dfsg/.pybuild/pythonX.Y_3.5/build/tests/test_symlink_time_handling/Snakefile", line 37, in <module>^[[0m
Removing output files of failed job make_output since they might be corrupted:
output_link
...


might be a sign that the test runs for Python2 and Python3 and the
temporary directory is not cleaned up.  I'll leave the task to sort
this out for the next upload to Kevin and uploaded.

Kevin, one additional remark:  Please use the same ID in Uploaders
and as changelog owner.  I fixed this in d/changelog - otherwise it
is regarded as NMU.

Kind regards

     Andreas.

-- 
http://fam-tille.de