BETSE 0.9.2.0 bumped.

[leycec]

This commit bumps conda-forge hosting to the most recent stable release: BETSEE 0.9.2.0 (Luckiest Levin). The PyPI-hosted domain of this package's tarball has been updated from the now seemingly dead pypi.io to pypi.org. This is what redirects are for, infrastructure maintainers. </sigh>

@conda-forge-admin, please rerender

Checklist

• Used a fork of the feedstock to propose changes
• Bumped the build number (if the version is unchanged)
• Reset the build number to 0 (if the version changed)
• Re-rendered with the latest conda-smithy (Use the phrase @conda-forge-admin, please rerender in a comment in this PR for automated rerendering)
• Ensured the license file is being packaged.

[conda-forge-linter]

Hi! This is the friendly automated conda-forge-linting service.

I just wanted to let you know that I linted all conda-recipes in your PR (recipe) and found it was in an excellent condition.

[leycec]

@conda-forge-admin, please rerender

[conda-forge-linter]

Hi! This is the friendly automated conda-forge-webservice.

I tried to re-render for you, but it looks like there was nothing to do.

[conda-forge-linter]

Hi! This is the friendly automated conda-forge-linting service.

I just wanted to let you know that I linted all conda-recipes in your PR (recipe) and found it was in an excellent condition.

[leycec]

@conda-forge-admin, please rerender

[conda-forge-linter]

Hi! This is the friendly automated conda-forge-webservice.

I tried to re-render for you, but it looks like there was nothing to do.

[leycec]

Urgh. CircleCI now spontaneously hiccups up the infamous libGL.so.1 import error:

+ betsee --help
Traceback (most recent call last):
  File "/home/conda/feedstock_root/build_artifacts/betsee_1548409688303/_test_env_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_p/bin/betsee", line 10, in <module>
    sys.exit(main())
  File "/home/conda/feedstock_root/build_artifacts/betsee_1548409688303/_test_env_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_p/lib/python3.7/site-packages/betsee/__main__.py", line 93, in main
    from betsee.guimetaapp import BetseeMetaApp
  File "/home/conda/feedstock_root/build_artifacts/betsee_1548409688303/_test_env_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_p/lib/python3.7/site-packages/betsee/guimetaapp.py", line 24, in <module>
    from betsee.lib.pyside2.cache.guipsdcache import CachePolicy
  File "/home/conda/feedstock_root/build_artifacts/betsee_1548409688303/_test_env_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_p/lib/python3.7/site-packages/betsee/lib/pyside2/cache/guipsdcache.py", line 23, in <module>
    from betsee.gui.simconf.stack.widget.guisimconfradiobtn import (
  File "/home/conda/feedstock_root/build_artifacts/betsee_1548409688303/_test_env_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_p/lib/python3.7/site-packages/betsee/gui/simconf/stack/widget/guisimconfradiobtn.py", line 21, in <module>
    from PySide2.QtWidgets import QButtonGroup, QRadioButton
ImportError: libGL.so.1: cannot open shared object file: No such file or directory
Tests failed for betsee-0.9.2.0-py_0.tar.bz2 - moving package to /home/conda/feedstock_root/build_artifacts/broken

If I had fine-grained control over the CircleCI configuration deployed here (...do I?), I'd just manually install the libgl1 system library (e.g., with apt install libgl1-mesa-glx from an Anaconda docker image). I note the relevant .circleci and .ci_support dot directories, of course, but suspect that any desperate hacks I stuff into those directories would simply be overwritten by the next automated conda-forge rerendering run. Is there a sane(r) way to go about this?

@mingwandroid and @CJ-Wright: I summon thee! You probably have more fulfilling quests to dispatch on a wintry Friday's eve, but the default CircleCI configuration for Qt-based conda-forge feedstocks appears to have gone badly awry. Although I know nothing, a recently misconfigured docker image might be the culprit.

Apologies for bothering you both so late into a bone-chilling January TGIF. May the underlying bug(s) be trivially dispatched and your weekends be merry and bright.

[CJ-Wright]

I think you'll find this thread useful, or at least informative.
https://gitter.im/conda-forge/conda-forge.github.io?at=5c409b109bfa375aab171e9b

[leycec]

Awesome. Thanks for the heads up! I should have thought to check the gitter first. Because I am dense and only slightly inebriated, I'm still unclear what needs doing here. Maybe the PySide2 feedstock needs a yum_requirements.txt explicitly listing libgl1 as a test-time requirement?

Let me pore through @mingwandroid's wizardly responses more deeply. 🤔

[leycec]

Ah. Jetz ist alles klar. @isuruf correctly identified the core issue – which, as expected, was an accidental docker misstep:

We didn't have this problem because linux-anvil had mesa-libGL installed, but linux-anvil-comp7 doesn't

..and @mariusvniekerk suggested the obvious solution – which, as expected, was amending the current dockerfile to install mesa-libGL:

probably worthwhile to update the docker images to just have that in?

[leycec]

Addendum: a relevant year-old issue already exists at "add libgl to Anaconda docker images." Of course, the last comment on that issue was... nearly a year ago. So this is why we can't have good things. 🤕

[leycec]

@conda-forge-admin, please rerender

[conda-forge-linter]

Hi! This is the friendly automated conda-forge-webservice.

I tried to re-render for you, but it looks like there was nothing to do.

[leycec]

This will probably simply fail again, but let's force another CircleCI rerun. If this does fail again, we'll have little choice but to arthritically shake our fist at the putrescent winter sky and:

1. Temporarily disable all tests.
2. Accept this pull request.
3. Wait until packages are built.
4. Submit a new pull request re-enabling tests.

y u do this 2 me, conda-forge.

[conda-forge-linter]

Hi! This is the friendly automated conda-forge-linting service.

I just wanted to let you know that I linted all conda-recipes in your PR (recipe) and found it was in an excellent condition.

[leycec]

Very well. You have chosen the bazooka, CircleCI. 💥

[leycec]

Let's see if we remember to re-enable functional tests in several months.

Since this glass is clearly half-empty, my gambling money is on "Nope."