FTBFS on all recent uploads

[Vagrant Cascadian]

On 2023-07-27, Peter Blackman wrote:
> I can see qt6ct has been rescheduled. Thanks.
>
> Looks OK on my packages overview, and the tracker,
> but is actually unreproducible!

This is because those only track weather trixie/testing is reproducible,
and unstable and experimental test more variations.


> FWIW, the note looks wrong.
> https://tests.reproducible-builds.org/debian/rb-pkg/unstable/amd64/qt6ct.html
>
>
> The note says
>      timestamps_in_source_generated_by_rcc

Yeah, I suspect quite a few of that particular issue are misfiled,
because many of them consistently build fine in trixie, just not
unstable...

Compare the unstable results to trixie:

  https://tests.reproducible-builds.org/debian/issues/unstable/timestamps_in_source_generated_by_rcc_issue.html

  5 FTBS, 66 unreproducible

  https://tests.reproducible-builds.org/debian/issues/trixie/timestamps_in_source_generated_by_rcc_issue.html

  49 reproducible, 2 FTBFS, 18 unreproducible

So my hunch is only ~18 of those are correctly marked, as timestamps
variations are more-or-less always tested on all suites.


> but the logs show the problem to be with NT_GNU_BUILD_ID.
> Usually an rpath issue.
> https://reproducible-builds.org/docs/deterministic-build-systems/#cmake-notes

Yeah, probably something triggered by build paths, which is not varied
in trixie/testing. The typical cmake rpath issue is fairly easy to
verify, by passing flags to configure:

  https://tests.reproducible-builds.org/debian/issues/unstable/cmake_rpath_contains_build_path_issue.html

Although sometimes there may be other causes.

Thanks for looking into it, hope that gives some good leads!

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/20230727/6c666588/attachment.sig>