[BUG] - Cannot install from conda

[dodu94]

Describe the bug

I tried to install geouned from conda on a fresh enviroment (python 3.11) without success. Error can be seen in the screenshot.

To Reproduce

Steps to reproduce the behavior: for example:

1. create a fresh conda environment
2. follow installation guide for conda and windows
3. See error

Screenshots

Please complete the following information):

• OS: Windows
• GEOUNED version: latest conda package?
• Python version: 3.11

[dodu94]

different PC, newer version of conda (22.X instead of an old 4.X) and I get a different error:

[shimwell]

Looks like the first error was present in an old version of conda and fixed in 4.6.0 https://github.com/conda/conda/issues/7776

[shimwell]

Looks like the 2nd error can be fixed by adding to the windows paths

See this comment https://github.com/conda/conda/issues/8836#issuecomment-590282285

[dodu94]

Is it worth to start a troubleshooting section in the doc?

[shimwell]

Up to you, I think these are conda issues instead of geouned issues.

I suspect they don't occur when using miniforge instead of anaconda or when using mamba instead of conda

[SterlingButters]

Working for me in conda 23.3.1

[akolsek]

@dodu94, whenever you can, please confirm if the error persists with the new release on conda-forge.

[dodu94]

I still get the same error for conda 22.X (on my laptop).

On conda 23.X (on my WS), instead, it took something like 20 mins if this endless looping:

Collecting package metadata (current_repodata.json): done
Solving environment: unsuccessful initial attempt using frozen solve. Retrying with flexible solve.
Solving environment: unsuccessful attempt using repodata from current_repodata.json, retrying with next repodata source.
Collecting package metadata (repodata.json): done
Solving environment: |

but finally it managed to install

[shimwell]

Conda appears to on version v24.7.1 at the moment, would it be worth trying a newer version

we could add a hint to update the conda version to the install docs if that helps?

For speed of solve then I think mamba is faster option

[dodu94]

Also please consider the fact that conda now moved to a mandatory paid plan for organizations of more than 200 employees (https://www.anaconda.com/pricing) so its popularity may fall a bit in the future (at least in pure open source environments)

[shimwell]

Yes a pip install would be great. I thought the use of miniforge + mamba + conda-forge helps avoid the paid plan