Skip to content

Comment on Don't use the mailing lists

Comments

Ultimately open source software is approved by many eyes, for which Github seems like a natural place. Many excellent packages are already only available via devtools::install_github(). CRAN is nice and easy (especially for people starting with R or with programming) but is not a necessity.

As a user I'm a bit wary of installing packages directly from GitHub. The CRAN review process isn't perfect, but it's much less likely that some malicious code accidentally slipped in, or that someone is namesquatting on a popular package name but is not the genuine package. With a random GitHub link I'd feel like I need to do this reviewing work every time myself: investigate who owns the GitHub repository, what their track record is, what recent commits there are and by whom and how reviewed, etc. That's a lot more work than using the heuristic that "CRAN code is probably safe". More mundanely, CRAN typically does a good job with dependencies, and is usually in a working state.

I could see alternate curated repositories from CRAN, but to switch from CRAN I'd still want some kind of curation.

AboutSource Built by g1lg1l

Hackerly is an independent reader for Hacker News, built on the public HN API. Not affiliated with Y Combinator.