Poetry does not output values of env markers under which a version conflict occurs

The following bug is reproducible on both latest stable (1.1.13) and pre-release (1.2.0b1) versions of poetry.

Problem

Given a simple pyproject.toml, containing only:

[tool.poetry]
name = "meta"
version = "0.0.1"
description = "No code, just dependencies."
authors = ["Tomas Babej <tomas@test.com>"]
license = "MIT"


[tool.poetry.dependencies]
python = "3.9.12"
numpy = "1.19.5"
pandas = "1.4.1"

poetry update fails with:

  SolverProblemError

  Because pandas (1.4.1) depends on numpy (>=1.20.0)
   and meta depends on numpy (1.19.5), pandas is forbidden.

The pineed versions of numpy and pandas are, however, compatible on my platform machine (pandas requires >=1.18.5), as evidenced by installing these dependencies using pip into the same virtual environment:

pip3 install numpy==1.19.5 pandas==1.4.1
Looking in indexes: https://pypi.org/simple, https://pypi.proteinqure.com/
Collecting numpy==1.19.5
  Downloading numpy-1.19.5-cp39-cp39-manylinux2010_x86_64.whl (14.9 MB)
     |████████████████████████████████| 14.9 MB 5.0 MB/s            
Collecting pandas==1.4.1
  Using cached pandas-1.4.1-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (11.7 MB)
Requirement already satisfied: pytz>=2020.1 in /home/tbabej/.cache/pypoetry/virtualenvs/meta-WEj54WUm-py3.9/lib/python3.9/site-packages (from pandas==1.4.1) (2022.1)
Requirement already satisfied: python-dateutil>=2.8.1 in /home/tbabej/.cache/pypoetry/virtualenvs/meta-WEj54WUm-py3.9/lib/python3.9/site-packages (from pandas==1.4.1) (2.8.2)
Requirement already satisfied: six>=1.5 in /home/tbabej/.cache/pypoetry/virtualenvs/meta-WEj54WUm-py3.9/lib/python3.9/site-packages (from python-dateutil>=2.8.1->pandas==1.4.1) (1.16.0)
Installing collected packages: numpy, pandas
  Attempting uninstall: numpy
    Found existing installation: numpy 1.19.2
    Uninstalling numpy-1.19.2:
      Successfully uninstalled numpy-1.19.2
  Attempting uninstall: pandas
    Found existing installation: pandas 1.4.0
    Uninstalling pandas-1.4.0:
      Successfully uninstalled pandas-1.4.0
Successfully installed numpy-1.19.5 pandas-1.4.1

Further investigation

Verification that python 3.9 wheel for pandas 1.4.1 allows numpy 1.19.5

Downloading the wheel for numpy 1.19.5 on Python3.9 (python3.9 -m pip download pandas==1.4.1), one can see the following dependencies in the METADATA file:

Requires-Dist: python-dateutil (>=2.8.1)
Requires-Dist: pytz (>=2020.1)
Requires-Dist: numpy (>=1.18.5) ; platform_machine != "aarch64" and platform_machine != "arm64" and python_version < "3.10"
Requires-Dist: numpy (>=1.19.2) ; platform_machine == "aarch64" and python_version < "3.10"
Requires-Dist: numpy (>=1.20.0) ; platform_machine == "arm64" and python_version < "3.10"
Requires-Dist: numpy (>=1.21.0) ; python_version >= "3.10"

so this seems to be related to handling of markers. The platform_machine value for my workstation is:

>>> import platform
>>> platform.machine()
'x86_64'

so anything above 1.18.5 is a valid dependency on this platform.

Hints as to what is happening under the hood

Running poetry update with increased verbosity reveals the following:

$ poetry update -vvv
Using virtualenv: /home/tbabej/.cache/pypoetry/virtualenvs/meta-WEj54WUm-py3.9
Updating dependencies
Resolving dependencies...
   1: fact: meta is 0.0.1
   1: derived: meta
   1: fact: meta depends on numpy (1.19.5)
   1: fact: meta depends on pandas (1.4.1)
   1: selecting meta (0.0.1)
   1: derived: pandas (==1.4.1)
   1: derived: numpy (==1.19.5)
PyPI: 1 packages found for pandas 1.4.1
PyPI: Getting info for pandas (1.4.1) from PyPI
   0: Duplicate dependencies for numpy
   0: Different requirements found for numpy (>=1.18.5) with markers platform_machine != "aarch64" and platform_machine != "arm64" and python_version < "3.10", numpy (>=1.19.2) with markers platform_machine == "aarch64" and python_version < "3.10" and numpy (>=1.20.0) with markers platform_machine == "arm64" and python_version < "3.10".
   1: Version solving took 0.153 seconds.
   1: Tried 1 solutions.
   0: Retrying dependency resolution with the following overrides ({Package('pandas', '1.4.1'): {'numpy': <Dependency numpy (>=1.18.5)>}}).
   1: fact: meta is 0.0.1
   1: derived: meta
   1: fact: meta depends on numpy (1.19.5)
   1: fact: meta depends on pandas (1.4.1)
   1: selecting meta (0.0.1)
   1: derived: pandas (==1.4.1)
   1: derived: numpy (==1.19.5)
   1: fact: pandas (1.4.1) depends on python-dateutil (>=2.8.1)
   1: fact: pandas (1.4.1) depends on pytz (>=2020.1)
   1: fact: pandas (1.4.1) depends on numpy (>=1.18.5)
   1: selecting pandas (1.4.1)
   1: derived: pytz (>=2020.1)
   1: derived: python-dateutil (>=2.8.1)
PyPI: No release information found for numpy-0.9.6, skipping
PyPI: No release information found for numpy-0.9.8, skipping
PyPI: No release information found for numpy-1.0.3, skipping
PyPI: No release information found for numpy-1.0.4, skipping
PyPI: No release information found for numpy-1.0b1, skipping
PyPI: No release information found for numpy-1.0b4, skipping
PyPI: No release information found for numpy-1.0b5, skipping
PyPI: No release information found for numpy-1.0rc1, skipping
PyPI: No release information found for numpy-1.0rc2, skipping
PyPI: No release information found for numpy-1.0rc3, skipping
PyPI: No release information found for numpy-1.1.1, skipping
PyPI: No release information found for numpy-1.2.0, skipping
PyPI: No release information found for numpy-1.2.1, skipping
PyPI: No release information found for numpy-1.4.0, skipping
PyPI: 1 packages found for numpy 1.19.5
PyPI: Getting info for numpy (1.19.5) from PyPI
PyPI: No dependencies found, downloading archives
PyPI: Downloading sdist: numpy-1.19.5.zip
   1: selecting numpy (1.19.5)
PyPI: 6 packages found for pytz >=2020.1
PyPI: Getting info for pytz (2022.1) from PyPI
PyPI: No dependencies found, downloading archives
PyPI: Downloading wheel: pytz-2022.1-py2.py3-none-any.whl
   1: selecting pytz (2022.1)
PyPI: No release information found for python-dateutil-0.1, skipping
PyPI: No release information found for python-dateutil-0.3, skipping
PyPI: No release information found for python-dateutil-0.4, skipping
PyPI: No release information found for python-dateutil-0.5, skipping
PyPI: No release information found for python-dateutil-1.0, skipping
PyPI: No release information found for python-dateutil-1.1, skipping
PyPI: No release information found for python-dateutil-1.2, skipping
PyPI: No release information found for python-dateutil-2.0, skipping
PyPI: 2 packages found for python-dateutil >=2.8.1
PyPI: Getting info for python-dateutil (2.8.2) from PyPI
   1: fact: python-dateutil (2.8.2) depends on six (>=1.5)
   1: selecting python-dateutil (2.8.2)
   1: derived: six (>=1.5)
PyPI: 18 packages found for six >=1.5
PyPI: Getting info for six (1.16.0) from PyPI
PyPI: No dependencies found, downloading archives
PyPI: Downloading wheel: six-1.16.0-py2.py3-none-any.whl
   1: selecting six (1.16.0)
   1: Version solving took 15.523 seconds.
   1: Tried 1 solutions.
   0: Retrying dependency resolution with the following overrides ({Package('pandas', '1.4.1'): {'numpy': <Dependency numpy (>=1.19.2)>}}).
   1: fact: meta is 0.0.1
   1: derived: meta
   1: fact: meta depends on numpy (1.19.5)
   1: fact: meta depends on pandas (1.4.1)
   1: selecting meta (0.0.1)
   1: derived: pandas (==1.4.1)
   1: derived: numpy (==1.19.5)
   1: fact: pandas (1.4.1) depends on python-dateutil (>=2.8.1)
   1: fact: pandas (1.4.1) depends on pytz (>=2020.1)
   1: fact: pandas (1.4.1) depends on numpy (>=1.19.2)
   1: selecting pandas (1.4.1)
   1: derived: pytz (>=2020.1)
   1: derived: python-dateutil (>=2.8.1)
   1: selecting numpy (1.19.5)
   1: selecting pytz (2022.1)
   1: fact: python-dateutil (2.8.2) depends on six (>=1.5)
   1: selecting python-dateutil (2.8.2)
   1: derived: six (>=1.5)
   1: selecting six (1.16.0)
   1: Version solving took 0.007 seconds.
   1: Tried 1 solutions.

This looks like a successful version resolution, but the resolution in fact attempts to continue further and eventually fails

...
   0: Retrying dependency resolution with the following overrides ({Package('pandas', '1.4.1'): {'numpy': <Dependency numpy (>=1.20.0)>}}).
   1: fact: meta is 0.0.1
   1: derived: meta
   1: fact: meta depends on numpy (1.19.5)
   1: fact: meta depends on pandas (1.4.1)
   1: selecting meta (0.0.1)
   1: derived: pandas (==1.4.1)
   1: derived: numpy (==1.19.5)
   1: fact: pandas (1.4.1) depends on python-dateutil (>=2.8.1)
   1: fact: pandas (1.4.1) depends on pytz (>=2020.1)
   1: fact: pandas (1.4.1) depends on numpy (>=1.20.0)
   1: conflict: pandas (1.4.1) depends on numpy (>=1.20.0)
   1: ! not  numpy (>=1.20.0) is satisfied by  numpy (1.19.5)
   1: ! which is caused by "meta depends on numpy (1.19.5)"
   1: ! thus: pandas is forbidden
   1: !  pandas (1.4.1) is satisfied by  pandas (1.4.1)
   1: ! which is caused by "meta depends on pandas (1.4.1)"
   1: ! thus: version solving failed
   1: Version solving took 0.004 seconds.
   1: Tried 1 solutions.

Reproduce the issue with the shown pyproject.toml and poetry update -vvv, then follow the dependency-resolution trace around the pandas and numpy environment markers. Done means compatible platform-specific requirements resolve correctly and any remaining conflict identifies the relevant marker values.