Cloning reproducible-builds environment

[Jeff]

Hi Chris,

On 08/01/2020 00:57, Chris Lamb wrote:
> But are you claiming that you…
> 
>>>> Is this a bug, then, that tiff2pdf is ignoring the TZ environmental
>>>> variable?
> 
> … believe that you have fixed this ^ bug? If so, it would be
> incurious of me if I did not request a link to said change… :)

I didn't express myself very well. I found a bug in my code causing it
not to parse the GMT-14 timezone, and fixed it here:

https://sourceforge.net/p/gscan2pdf/code/ci/3fa16ee77c5fffde5b8c48a0213d387b087fd2e0/

That fixed the general problem, and just leaves me scratching my head
over this:

https://sourceforge.net/p/gscan2pdf/code/ci/master/tree/t/1628_import_pdf_metadata.t

The above skipped test fails in reprotest (but not in any normal
schroot). Whatever variants I change like this:

reprotest -vv --variations="-time" 'prove -lv
t/1628_import_pdf_metadata.t' 'dist/*.tar.gz'

The build log still gives me:

INFO:reprotest:build "control": FIX environment, FIX build_path, FIX
kernel, FIX aslr, FIX num_cpus, FIX time, FIX user_group, FIX
fileordering, FIX domain_host, FIX home, FIX locales, FIX exec_path, FIX
timezone, FIX umask

How do I narrow down which variant caused the problem?

How do I reproduce it outside reprotest?

Regards

Jeff