installation fails - channel settings

[ThomasRoosli]

On a virtual Linux Machine the advanced installation failed whilst installing the requirements file from petals. With the following message:

Could not solve for environment specs
The following packages are incompatible
└─ scikit-image >=0.22  is installable with the potential options
   ├─ scikit-image 0.22.0 would require
   │  └─ tifffile >=2022.8.12  with the potential options
   │     ├─ tifffile [2022.10.10|2022.8.12] would require
   │     │  └─ imagecodecs >=2021.11.20  with the potential options
   │     │     ├─ imagecodecs [2021.11.20|2022.12.22|...|2023.8.12] would require
   │     │     │  ├─ libbrotlicommon >=1.0.9,<1.1.0a0 , which can be installed;
   │     │     │  └─ libbrotlienc >=1.0.9,<1.1.0a0  with the potential options
   │     │     │     ├─ libbrotlienc 1.0.9 would require
   │     │     │     │  └─ libbrotlicommon 1.0.9 h166bdaf_7, which can be installed;
   │     │     │     ├─ libbrotlienc 1.0.9 would require
   │     │     │     │  └─ libbrotlicommon 1.0.9 h166bdaf_8, which can be installed;
   │     │     │     ├─ libbrotlienc 1.0.9 would require
   │     │     │     │  └─ libbrotlicommon 1.0.9 h7f98852_5, which can be installed;
   │     │     │     ├─ libbrotlienc 1.0.9 would require
   │     │     │     │  └─ libbrotlicommon 1.0.9 h7f98852_6, which can be installed;
   │     │     │     ├─ libbrotlienc 1.0.9 would require
   │     │     │     │  └─ libbrotlicommon 1.0.9 h166bdaf_9, which can be installed;
   │     │     │     └─ libbrotlienc 1.0.9 would require
   │     │     │        └─ libbrotlicommon 1.0.9 h5eee18b_7, which can be installed;
   │     │     ├─ imagecodecs [2022.12.22|2022.12.24|...|2023.9.4] would require
   │     │     │  └─ python >=3.11,<3.12.0a0 , which can be installed;
   │     │     ├─ imagecodecs [2021.11.20|2022.12.22|...|2023.9.4] would require
   │     │     │  └─ python >=3.9,<3.10.0a0 , which can be installed;
   │     │     ├─ imagecodecs [2023.9.18|2023.9.4] would require
   │     │     │  └─ libjpeg-turbo [>=2.1.5.1,<3.0a0 |>=3.0.0,<4.0a0 ], which requires
   │     │     │     └─ jpeg <0.0.0a , which can be installed;
   │     │     ├─ imagecodecs 2023.9.18 would require
   │     │     │  └─ python >=3.12.0rc3,<3.13.0a0  with the potential options
   │     │     │     ├─ python 3.12.0, which can be installed;
   │     │     │     └─ python 3.12.0rc3 would require
   │     │     │        └─ _python_rc  , which does not exist (perhaps a missing channel);
   │     │     ├─ imagecodecs [2021.11.20|2022.12.22|...|2023.1.23] would require
   │     │     │  └─ python >=3.8,<3.9.0a0 , which can be installed;
   │     │     ├─ imagecodecs 2023.1.23 would require
   │     │     │  └─ python >=3.12,<3.13.0a0 , which can be installed;
   │     │     └─ imagecodecs 2021.11.20 would require
   │     │        └─ python >=3.7,<3.8.0a0 , which can be installed;
   │     ├─ tifffile [2023.1.23.1|2023.2.2|...|2023.8.12] would require
   │     │  └─ imagecodecs >=2023.1.23 , which can be installed (as previously explained);
   │     ├─ tifffile [2023.12.9|2023.8.12|...|2023.9.26] would require
   │     │  └─ imagecodecs >=2023.8.12 , which can be installed (as previously explained);
   │     ├─ tifffile 2023.4.12 would require
   │     │  └─ python >=3.11,<3.12.0a0 , which can be installed;
   │     ├─ tifffile 2023.4.12 would require
   │     │  └─ python >=3.12,<3.13.0a0 , which can be installed;
   │     ├─ tifffile 2023.4.12 would require
   │     │  └─ python >=3.8,<3.9.0a0 , which can be installed;
   │     └─ tifffile 2023.4.12 would require
   │        └─ python >=3.9,<3.10.0a0 , which can be installed;
   ├─ scikit-image 0.22.0 would require
   │  └─ python >=3.11,<3.12.0a0 , which can be installed;
   ├─ scikit-image 0.22.0 would require
   │  └─ python >=3.12,<3.13.0a0 , which can be installed;
   └─ scikit-image 0.22.0 would require
      └─ python >=3.9,<3.10.0a0 , which can be installed.

After many trials, I found that mamba does not support the defaults channel: https://mamba.readthedocs.io/en/latest/user_guide/troubleshooting.html#mixing-the-defaults-and-conda-forge-channels
According to Mamba we should change our requirements.yml in python and petals to:

channels:
  - conda-forge
  - nodefaults

After this change I was able to successfully install CLIMADA python and petals using the advanced installation instructions.

[emanuel-schmid]

Thanks for bringing this up!

[peanutfun]

@ThomasRoosli can you give this another try with the latest develop branch?

[peanutfun]

Channel settings have been adjusted in Core and Petals. @ThomasRoosli, please reopen this issue if problems persist.