Cloning reproducible-builds environment

[Vagrant Cascadian]

On 2019-12-19, Jeff wrote:
> On 18/12/2019 22:47, Holger Levsen wrote:
>> what version of reprotest were you using? (newer versions dont use
>> faketime for the 2nd build, which likely has been causing your
>> problems.)
>
> The 2nd build fails on the official builds, too:
>
> https://tests.reproducible-builds.org/debian/logs/unstable/amd64/gscan2pdf_2.6.2-1.build2.log.gz
>
> If you could show me how to run the second build first, I'll have a
> fighting change of fixing the problem in my package.
>
> Thanks for your work on reproducible builds!

The variations done on the first and second builds are documented here:

  https://tests.reproducible-builds.org/debian/index_variations.html

You could run a build manually with some of the easy ones changed, such
as locale (the second build runs with a UTF-8 locale, the first in
non-utf8 C).


According to your log, it does appear to be time/date/timezone related:

#   Failed test 'datetime - timezone'
#   at t/1628_import_pdf_metadata.t line 46.
#     Structures begin differing at:
#          $got->[0] = '2019'
#     $expected->[0] = '2018'
# Looks like you failed 1 test of 6.
t/1628_import_pdf_metadata.t .................. 
1..6
ok 1 # skip skipping the full datetime + timezone test until I understand it better
not ok 2 - datetime - timezone
ok 3 - author
ok 4 - subject
ok 5 - keywords
ok 6 - title
Dubious, test returned 1 (wstat 256, 0x100)
Failed 1/6 subtests 
	(less 1 skipped subtest: 4 okay)

...

Test Summary Report
-------------------
t/1628_import_pdf_metadata.t                (Wstat: 256 Tests: 6 Failed: 1)
  Failed test:  2
  Non-zero exit status: 1
Files=232, Tests=1129, 407 wallclock secs ( 0.92 usr  0.96 sys + 247.45 cusr 46.72 csys = 296.05 CPU)
Result: FAIL


As an alternate approach, you could run with reprotest and disable the
test suites, since the package seems to build otherwise, to produce
something that diffoscope can at least compare.

Then if you fix the issues leading to the differences, you might be
lucky enough that it also fixes the issues in the test suite. :)


Good luck!


live well,
  vagrant
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 227 bytes
Desc: not available
URL: <http://alioth-lists.debian.net/pipermail/reproducible-builds/attachments/20191219/b731ec54/attachment.sig>