Giters

rstudio / rstudio-docker-products

Docker images for RStudio Professional Products

Home Page:https://hub.docker.com/u/rstudio

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Update primary/symlinked R version to 4.2.2 for the base images

tnederlof opened this issue · comments

commented

Currently, base images are being built using R_VERSION of 3.6.3 which is also the symlinked R version (causing it to be the default R version in Pro Products further upstream).

Since 3.6.3 is over three years old it would be more appropriate to use anR_VERSION of 4.2.2 instead in base images. 3.6.3 could be R_VERSION_ALT although we may want to consider a more modern version such as 4.1.3

Variables set in the CI here:
https://github.com/rstudio/rstudio-docker-products/blob/main/ci.Justfile

commented

Is there a generalizable heuristic that Posit might consider using to define which versions of R are installed in the base image?

For example, there are differences across source within r-session-complete:

  1. https://github.com/rstudio/rstudio-docker-products/blob/main/r-session-complete/Justfile#L15-L16

    • R_VERSION := "4.1.0"
    • R_VERSION_ALT := "3.6.2"

  2. https://github.com/rstudio/rstudio-docker-products/blob/main/r-session-complete/Dockerfile.ubuntu1804#L1-L2

    • R_VERSION := "3.6.2"
    • R_VERSION_ALT := "4.1.0"

E.g., an approach we've been using:

  • R_VERSION: Latest patch of the 2nd-to-latest minor release: 4.1.3
  • R_VERSION_ALT: Latest patch of the Nth-to-latest minor release: e.g. 3.6.3
commented

Thanks for bringing this up- we are talking about this internally and trying to determine the best way to do this. There's some complications around making this change given it would be breaking. We are trying to think through how this process would work, as it has been a bit complicated to do things like this in the past. @ianpittwood and @colearendt are currently looking to figure this out.