Skip to content

Conversation

@aster-void
Copy link
Contributor

@aster-void aster-void commented Oct 29, 2025

closes #20

New features

  • FOD oriented fetcher, so user only has to specify hash once.
  • in-place lockfile specification { npmLock = ./package-lock.json; }, rather than builder-pattern
    • old builder pattern styles are still supported.

Code refactors without behavior changes

  • separate lib/parsers.nix and lib/legacyFetcher.nix from main prisma.nix

Breaking changes

I believe there are no breaking changes, so old systems should continue to work without any code changes.

@VanCoding VanCoding merged commit 1e5e45c into VanCoding:main Oct 31, 2025
4 checks passed
@VanCoding
Copy link
Owner

This is suuper nice! Thank you ❤️

I wonder, should we print a message to users of the legacy API that there's a new one and that the old one is deprecated?
If we deprecate the old API, we can even remove its documentation from the readme and instead link to an older version of the readme.

What do you think?

@aster-void aster-void deleted the feat/new-fetcher branch October 31, 2025 15:57
@aster-void
Copy link
Contributor Author

Agreed, that will make the README much cleaner 👍

@VanCoding
Copy link
Owner

btw, I noticed that the new API does not work yet as described in the readme. in the "old-ish" way with for example fromNpmLock it still works, but with the "npmLock" property it doesn't. Maybe it's just me using it wrong, though. But the tests also don't conver the new way yet if I understand it correctly.

@VanCoding
Copy link
Owner

Agreed, that will make the README much cleaner 👍

Hmm, or on a second thought, maybe we can also move it to another file so it's still around. This could also serve as documentation for us, we need to support the old way for a bit after all.

@VanCoding
Copy link
Owner

I've made a fix for using it with npmLock=..., in 0d2d096. I hope this makes sense.

@aster-void
Copy link
Contributor Author

btw, I noticed that the new API does not work yet as described in the readme. in the "old-ish" way with for example fromNpmLock it still works, but with the "npmLock" property it doesn't. Maybe it's just me using it wrong, though. But the tests also don't conver the new way yet if I understand it correctly.

That's completely on me, I'm sorry. Thank you for the quick fix.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

Feature Request: Use Fixed Output Derivation to reduce required hash updates

2 participants