Musings on a monorepo versus developer oriented distribution

could help with a discourse topic for each proposal? I’m a bit lost with the proposals.

I’m struggling myself too, but let me try to shed some light onto the current discussion in this area:

I do think we’re starting to debate ‘the same’ topic in three different places now, or rather three so closely connected topics, that they heavily influence each other:

  • in this topic, we’re debating mono- vs multi-repo
  • in SIG Repos: How should they work? we’re already discussing how our multi-repo approach “will look like” (in the sense that ‘mono-repo’ isn’t an option that’s part of the options in that specific topic)
  • in On the future of our nixpkgs fork, we’re discussing how we will (not) integrate nixpkgs into our own package set(s); but with the base assumption also kind of being that we have a multi-repo approach

Nevertheless, you raise some excellent questions.

At some point, (part of) the answer was (or in my impression the answer was):

Let me try to summarize the current discussion and extract the main points still in discussion:

  • Goals (from user perspective):
    • SIG repos can be used individually, reducing size and eval time
    • top-level still provides what previously ‘Nixpkgs’ was - a one-stop location for all packages
  • Goals (from SIG/maintainer perspective):
    • reduce maintainer burden
    • scoped repos, with less noise
    • less moving parts in each SIG repo

I’m very unsure how to proceed here, with the discussion being spread across 3 different topics,
with wildly varying underlying assumptions - but also a lot of thoughts and discussion already spent on each of them.

1 Like