README

Workflow

1. "Prepare"
   1. Install R interpreter
   2. Add required repos to resolve syslibs
   3. Install pak
   4. Install awscli for remote s3 sync
2. "Build"
   1. For every package, do
      1. Clone from GitHub CRAN mirror
      2. Install all system dependencies via pak::local_install_dev_deps()
      3. Checkout each available tag
      4. Build binary package
      5. Create Archive/ directory
      6. Move all versions besides the most recent one to Archive/ (for source and binary)
      7. Create PACKAGES metadata file using cranlike package
      8. Store build metadata in sqlite DB
3. "Upload"
   1. Sync all binaries and sources to S3 bucket
4. -4-create-archive-rds.sh
   1. Create Meta/archive.rds file for each package so that remotes::install_version() can install archived versions
   2. Upload Meta/archive.rds to S3 bucket

Build metadata

The build metadata is stored in a sqlite database. The following tables are used:

• metadata contains the build metadata for each package:

package_name	tag	platform	error_occurred	build_timestamp

• metadata_summary contains aggregated build metadata for each package and platform:

package_name	platform	successful_builds	unsuccessful_builds	total_builds	percentage_successful_builds	average_build_time_per_tag

Support for archived versions

remotes::install_version()

remotes::install_version() is searching for a Meta/archive.rds file in the /src/contrib directory. This file must be a list of dataframes containing the information about the archived versions of the package.

Example:

con <- gzcon(url(sprintf("%s/src/contrib/Meta/archive.rds",
                           c("CRAN" = "https://cloud.r-project.org")), "rb"))
foo = readRDS(con)
foo[[1]]

pak::pak(package@version)

pak searches for Archive/<package> and is able to install all versions existing there. Ensure to use a clean cache if other repos have been used before, i.e. call pak::meta_clean(force = TRUE) if in doubt or when testing.

Platforms

The following platforms are supported:

• RHEL9

Lessons learned

• Newer R versions have issues compiling C code from older package Versions. Using an older R version to build the binaries is recommended.
• R 4.0.5 is used to build all binaries for R 4.x
• Binaries which cannot be built using R 4.0.5 often also require older system library or compiler versions and are out of scope for this project.