Cannot concretize Jupyter ecosystem, multiple versions of build deps with incompatible run deps #51134
Description
Steps to reproduce
Chasing down an issue trying to install py-geemap:
> spack spec py-geemap
==> Error: failed to concretize `py-geemap` for the following reasons:
1. Cannot satisfy 'py-jupyterlab@3'
required because py-plotly depends on py-jupyterlab@3 when @5:
2. Cannot satisfy 'py-jupyterlab@3'
required because py-ipytree depends on py-jupyterlab@3
3. Cannot satisfy 'py-jupyterlab@3'
required because py-ipyevents depends on py-jupyterlab@3
4. Cannot satisfy 'py-jupyterlab@3'
5. Cannot satisfy 'py-jupyterlab@4' and 'py-jupyterlab@3
required because py-jupyter-leaflet depends on py-jupyterlab@4
required because py-ipyleaflet depends on [email protected]
required because py-geemap depends on [email protected]:
required because py-geemap requested explicitly
required because py-plotly depends on py-jupyterlab@3 when @5:
6. Cannot satisfy 'py-jupyterlab@4' and 'py-jupyterlab@3
required because py-jupyter-leaflet depends on py-jupyterlab@4
required because py-ipyleaflet depends on [email protected]
required because py-geemap depends on [email protected]:
required because py-geemap requested explicitly
required because py-ipytree depends on py-jupyterlab@3
7. Cannot satisfy 'py-jupyterlab@4' and 'py-jupyterlab@3
required because py-jupyter-leaflet depends on py-jupyterlab@4
required because py-ipyleaflet depends on [email protected]
required because py-geemap depends on [email protected]:
required because py-geemap requested explicitly
required because py-ipyevents depends on py-jupyterlab@3 Organizing this more legibly:
- Cannot satisfy 'py-jupyterlab@4' and 'py-jupyterlab@3
- py-ipyevents depends on py-jupyterlab@3
- py-ipytree depends on py-jupyterlab@3
- py-jupyter-leaflet depends on py-jupyterlab@4
- py-plotly depends on py-jupyterlab@3
However, all 4 of these packages have a build-only dependency on py-jupyterlab. Furthermore, py-jupyterlab has:
tags = ["build-tools"]so I would expect the concretizer to allow multiple versions of py-jupyterlab.
My best guess is that another package in the DAG has a runtime dependency on py-jupyterlab and it isn't showing up in the error message. But I'm not sure how to debug this further. Pip doesn't have any issues with geemap/jupyterlab as each individual build environment is concretized separately from the overall run environment.
Error message
No response
Information on your system
- Spack: 1.1.0.dev0 (b187f87)
- Builtin repo: spack/spack-packages@381ecc4
- Python: 3.13.2
- Platform: darwin-sequoia-m2
Pinging @alalazo for the concretizer, @becker33 for the concretizer error messages, and @alecbcs because he reported this same issue in spack/spack-packages#1013
General information
- I have run
spack debug reportand reported the version of Spack/Python/Platform - I have searched the issues of this repo and believe this is not a duplicate
- I have run the failing commands in debug mode and reported the output
Metadata
Metadata
Assignees
Labels
Type
Projects
Status