Steps to reproduce

1. Clone and activate a new Conda-Analytics environment
2. Run conda install r-base r-arrow r-stringi (you can omit r-base, but the result is the same since the other packages both require it)
3. The installation will fail with a LibMambaUnsatisfiableError

I also tried the install command with --solver classic, but it also failed with Found conflicts! Looking for incompatible packages (it was still looking after about 2 hours, at which point I killed it).

Error details

The full error message is:

LibMambaUnsatisfiableError: Encountered problems while solving:
  - package r-arrow-10.0.1-r41hcb278e6_0 requires r-base >=4.1,<4.2.0a0, but none of the providers can be installed

Could not solve for environment specs
The following packages are incompatible
[extremely long tree of options omitted]

Pins seem to be involved in the conflict. Currently pinned specs:
 - sqlalchemy[version='<2.0']
 - pyspark=3.1.2
 - pyarrow=9.0.0
 - pandas[version='<2.0.0']
 - numpy[version='<1.24.0']
 - jupyterlab_server=2.25
 - jupyterlab=3.4.8
 - jupyterhub-systemdspawner=0.15.0
 - jupyterhub-ldapauthenticator=1.3.2
 - jupyterhub=1.5.0
 - jupyter_core=5.5
 - python=3.10

My full shell output, including the long option tree, is in P74369, but I recommend trying to reproduce it yourself since the terminal output color-codes the dependency tree according to what conflicts and what doesn't, which makes it a bit easier to understand.'

Workaround

In place of Arrow, Nanoparquet (conda install r-nanoparquet) can be used to load Parquet files. It isn't quite as heavy duty, but it's likely to work for most use cases. This make it possible to continue using Stringi and all the packages that depend on it.

As a result, while this is annoying, it has only moderate impact.