Confusion of installing R package in conda/mamba/miniforge3 R env in WSL

[Shijie-Jian]

System details:

Positron and OS details:

Positron Version: 2025.10.1 (system setup) build 4
Code - OSS Version: 1.103.0
Commit: b13dd1c
Date: 2025-10-14T22:13:39.363Z
Electron: 37.2.3
Chromium: 138.0.7204.100
Node.js: 22.17.0
V8: 13.8.500258-electron.0
OS: Windows_NT x64 10.0.19045

Session details:

R 4.4.3 with conda

Describe the issue:

I created a R virtual env by miniforge3 in WSL Ubuntu, and properly set "Positron > R: Custom Root Folders" with "~/miniforge3/envs" (miniforge3 envs directory), so I could select specific R interpreter session (R of conda).

However, I am not able to install R package from sources in CONSOLE, as it mentioned lack of "make" compilation tool (this tool actually existed in the R virtual env but not in base env created by miniforge3 by default, checked through "make --version" in these two envs).

Surprisedly, I could install the same R package from sources in TERMINAL in the activated R virtual env. After that the package could be "library or require" properly in TERMINAL.

I am not sure whether I missed some manual setting in Positron at this case.

[Shijie-Jian]

An additional information is these R packages were controlled by renv R package of each project.

[louhao1129]

My experience is identical to yours:
When using an R environment created by conda (miniforge3/mamba/anaconda), installing packages from source directly in the Positron R CONSOLE fails with errors like make: not found or similar messages about missing compilation tools (e.g., gcc).

However, if I open the TERMINAL in Positron manually run conda activate , and then install the package within the R session started in that terminal, everything works perfectly.

[juliasilge]

Yes, as I believe we've outlined elsewhere such as in #4398 (comment), we do not officially support conda currently and we do not believe that R package installation support works. If you want to experiment with using R installed via conda in Positron, you will need to install the packages via conda itself and not from the console.