[kixelated]

Follow-up to #5460, which merged with the no-go. This tested the maintainer's rename idea and filed the upstream issue. The verdict stays no-go until mozilla/uniffi-rs#3026 is resolved.

Rename probe

Same spike harness as #5460: crates net (namespace moq) and media linked into one libmoq_ffi. media's colliding types got distinct Rust names (MediaGroupConsumer, MediaFrame) so the bare-name TypeUniverse check passes. A per-language rename then mapped them back to GroupConsumer and Frame. UniFFI 0.32.2 has [bindings.{kotlin,swift,python}.rename] built in. The Go, Dart, and C++ forks have no rename option, so I patched each one locally (not pushed) to apply [bindings.<lang>.rename] through uniffi_bindgen::interface::rename, the same way the Kotlin backend does.

Generator	Renamed type and net type coexist	How the external type is referenced	Compiles and runs
Python (built in)	yes	module-qualified (moq.GroupConsumer)	yes, the test passes
Go (fork, patched)	yes	package-qualified (moq.GroupConsumer)	yes after fixing 2 lines by hand: uniffi's rename visitor skips the constructor's return type (*MediaGroupConsumer)
Kotlin (built in)	no	import uniffi.moq.GroupConsumer shadows the local class	no: media's own converters resolve to the net types
Swift (built in)	no	not imported; one-module layout	no: 'GroupConsumer' is ambiguous in one module. With one module per crate (plus import moq), the local type shadows the import (value type 'Frame' cannot have a stored property that recursively contains it)
Dart (fork, patched)	no	unqualified import "moq.dart", so the local type shadows it	generates and compiles, but media.GroupConsumer's constructor ends up typed against itself, so a consumer can't pass a net GroupConsumer
C++ (fork, patched)	no	sorts definitions by bare name	renaming the record panics (Cyclic dependency detected). Renaming only the object emits its converter into the net namespace (redefinition of moq::uniffi::FfiConverterGroupConsumer)
Generated C (fork)	n/a	n/a	the backend still refuses external types

Only the backends that refer to external types through their module (Python, Go) survive the rename. Fixing the rest means changing how each backend references external types, which is the upstream fix.

Upstream

I found no existing issue: the closest are mozilla/uniffi-rs#2732 (same crate, closed), mozilla/uniffi-rs#2153 (Swift/Python visibility), and mozilla/uniffi-rs#2985. So I filed mozilla/uniffi-rs#3026. It has a two-crate repro, the exact conflicting types error from TypeUniverse::add_known_type, a fix sketch (key by (module_path, name) and qualify cross-crate references per backend), and the Python silent drop. That drop looks fixed on uniffi main by mozilla/uniffi-rs#2985, which no 0.32.x release includes.

Change

quest/m1/ffi-shape/README.md records the rename result and links mozilla/uniffi-rs#3026 as the condition for revisiting the split.

Decisions

• ✅ Merge as is: the split stays no-go and the wrappers keep the namespaces (maintainer, 2026-10-13).
• ✅ Track Library mode: a type cannot reference an external type with the same name mozilla/uniffi-rs#3026 through the ffi-shape README link only, with no waiting quest (maintainer, 2026-10-13).

Public API and wire impact

None. Quest files only.

(Written by Claude Opus 5.5)

🤖 Generated with Claude Code

[chatgpt-codex-connector]

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review	Status	Commit	Review trigger
📝 Code Review	✅ Completed 2026-10-13T23:21:00.287659Z	ea704e7	Draft marked ready

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

[coderabbitai]

Warning

Review limit reached

You've used all free OSS reviews for now. Wait for the free limit to reset to keep reviewing this public repository.

Next included review available in 45 seconds.

Check out review usage here.

View limit details

Limit details: You’ve used all 4 included reviews currently available.

Learn how review limits work.

Review configuration:

⚙️ Run configuration

• Configuration used: Organization UI
• Review profile: CHILL
• Plan: Advanced
• Run ID: e1c9b62e-29b3-4e85-8da8-7a6dfa0a6d66

📥 Commits

Reviewing files that changed from the base of the PR and between 90776f9 and ea704e7.

📒 Files selected for processing: 1

• quest/m1/ffi-shape/README.md

• Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[kixelated]

Merge summary: this records the rename probe and links mozilla/uniffi-rs#3026 from quest/m1/ffi-shape/README.md. Quest files only, with no API or wire impact.

Decisions, confirmed by the maintainer:

• ✅ Merge as is. The per-layer crate split stays no-go and the wrappers keep the namespaces.
• ✅ Track Library mode: a type cannot reference an external type with the same name mozilla/uniffi-rs#3026 through the README link only, with no waiting quest.

CI is green (Check, Quest, Test). Codex gave a thumbs-up on head ea704e7. CodeRabbit didn't review because of its rate limit.

(Written by Claude Opus 5.5)