Does the Cargo team accept RFC: Packages as (optional) namespaces #3243

[epage]

rust-lang/rfcs#3243 has enough impact on cargo (and we want to make sure all voices are heard) that we wanted to have an FCP for this, rather than an informal yay/nay.

To avoid the "two team FCP" problem (both teams - 2 people must approve), we're holding the FCP here so its only "cargo team - 2 people"

Note: This issue exists purely for process purposes and feedback should be given on rust-lang/rfcs#3243

[epage]

@rfcbot fcp merge

Main effects on cargo

• Support :: in the middle of a package name, once
  • Including dep: feature syntax
• Ramifications for implicit feature names (do we support it for those?)
  • Decide during implementation/stabilization?
• What should the lib name be?
  • Decide during implementation/stabilization?
• Should we support renaming a package into a namespace?
  • Decide during implementation/stabilization?
• What should the generated .crate file names be?
• What should be used for :: in the index file paths?

This is assuming that namespace participation is only enforced within a registry when a package is published to it.

[rfcbot]

Team member @epage has proposed to merge this. The next step is review by the rest of the tagged team members:

• @Eh2406
• @Muscraft
• @arlosi
• @ehuss
• @epage
• @joshtriplett
• @weihanglo

No concerns currently listed.

Once a majority of reviewers approve (and at most 2 approvals are outstanding), this will enter its final comment period. If you spot a major issue that hasn't been raised at any point in this process, please speak up!

See this document for info about what commands tagged team members can give me.

[weihanglo]

What should the lib name be?

The current iteration of rust-lang/rfcs@482fae8 choose pkg as the autogenerated lib.name for namespace::pkg. This might introduce more name conflicts on build artifacts (see #6313). For example, a workspace contains both png::codec and avif::codec and want to build them into dylib.

This is not a blocker if postponing it to implementation stage is permitted, but it would be great if we can have a consensus on it.

[ehuss]

@weihanglo Can you clarify what kind of conflict you see? For rlibs, the filename should have a distinct hash between them.

The conflicts I can think of are:

• dylib/cdylib — these don't have hashes. This seems unlikely in practice, though something to consider.
• rustdoc — I'm going to leave a comment about this on the RFC.

[weihanglo]

Yes thanks, Eric. dylib/cdylib are things I am aware of. Might not be common, and if cargo can apply rename feature to namespaced package, everything would be okay. rustdoc's issue is more prominent.

[rfcbot]

🔔 This is now entering its final comment period, as per the review above. 🔔

[rfcbot]

The final comment period, with a disposition to merge, as per the review above, is now complete.

As the automated representative of the governance process, I would like to thank the author for their work and everyone else who contributed.

This will be merged soon.

[ehuss]

I'm going to close this as accepted. Thanks everyone!