[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Bug#1002344: hy: FTBFS: dh_auto_test: error: pybuild --test --test-pytest -i python{version} -p "3.10 3.9" returned exit code 13



On Wed, 22 Dec 2021 at 00:04, Lucas Nussbaum <lucas@debian.org> wrote:
> Source: hy
> Version: 0.19.0-2
> Severity: serious
> Justification: FTBFS
> Tags: bookworm sid ftbfs
> User: lucas@debian.org
> Usertags: ftbfs-20211220 ftbfs-bookworm
>
> Hi,
>
> During a rebuild of all packages in sid, your package failed to build
> on amd64.
> ...
> If you fail to reproduce this, please provide a build log and diff it with mine
> so that we can identify if something relevant changed in the meantime.

Ouch, thanks for the report.  I think the best fix for this is going
to be updating the package to the 1.0a3 upstream release (because I'm
fairly sure this is really about Hy 0.19 not actually supporting
Python 3.10 yet and the 1.0a3 release does), however that bump
includes the landmine of a complete documentation refactor that
creates a bit of a complex cycle with (the new)
https://github.com/hylang/sphinxcontrib-hydomain repository. 😞

I've played with building those together as a multiorig package (given
both the dependency cycle and that there aren't likely to be many/any
other consumers of "sphinxcontrib-hydomain" at this time), and ran
into a number of (hopefully minor?) problems, the first of which being
versioning (that repository has a tagged version, but it isn't
compatible with the latest Hy release nor the version of Sphinx in
Debian, but getting a commit that's good enough for the latter doesn't
fully resolve the former and requires further patching).

However, I'm definitely in a bit over my head here, so I'm adding
debian-python to CC hoping that perhaps someone there has some good
advice or ideas (maybe even a spare cycle or two). 🙈

My (non-functional) WIP can be found at
https://salsa.debian.org/hy-team/hy/-/compare/master...1.0a3 😇

If I bypass the Sphinx bits, I run into "INTERNALERROR> IndexError:
list index out of range" (and the full backtrace appears to all be in
non-Hy codelike pytest and "pluggy"), which I *think* is related to
funcparserlib needing to be 1.0.0a0+
(https://github.com/hylang/hy/pull/2164), but I am not certain.

♥,
- Tianon
  4096R / B42F 6819 007F 00F8 8E36  4FD4 036A 9C25 BF35 7DD


Reply to: