chore: re-pin cpp-sdk/qt-sdk/rust-sdk/protocol (full_api fixes) - #150
Conversation
Bumps the SDK inputs to the merged fixes surfaced by the full_api test chain: - logos-protocol -> 4775e63 (#21, qvariantToNlohmann container int-preservation) - logos-cpp-sdk -> 2f34804 (#103, bstr + composite-any codegen) - logos-qt-sdk -> 089142d (#10, QtProviderObject QVariantList/Map args) - logos-rust-sdk -> a55fdac (#26, clone any event params) Validated: all five test-modules full_api modules (C++/Rust providers, C++/Rust proxies, ui_qml plugin) build through this module-builder with no SDK overrides. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
There was a problem hiding this comment.
Pull request overview
Updates the Nix flake inputs/lockfile to consume merged full_api fixes in the Logos SDK ecosystem, so module-builder builds against the intended protocol + SDK revisions without per-module overrides.
Changes:
- Bump
logos-rust-sdkinput pin inflake.nixtoa55fdac…(full_api fix). - Update
flake.lockto new locked revs/hashes forlogos-cpp-sdk,logos-protocol,logos-qt-sdk, andlogos-rust-sdk.
Reviewed changes
Copilot reviewed 1 out of 2 changed files in this pull request and generated no comments.
| File | Description |
|---|---|
| flake.nix | Updates the pinned logos-rust-sdk input revision used by the builder. |
| flake.lock | Re-locks SDK/protocol dependencies to the specified full_api fix commits with updated narHash/lastModified metadata. |
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
📊 Doc-test reportsThe four ways to wrap a C library plus the two cross-language composition tours — each scaffolded into real modules, built against this commit, loaded in logoscore, and driven — rendered alongside the commands actually run and their output (updated each run, commit Pages can take a minute to update after the run finishes. |
…ecode logos-cpp-sdk 692af75 (#150) -- the generated cdylib dispatch bounds the argument count ABOVE as well as below, including for zero-parameter methods and the derived identity dispatch; and the lp surface's jsonToStdResult stops answering a non-object reply with a default StdLogosResult, which is byte-for-byte a provider refusal. logos-rust-sdk d3fe4b4 (#50) -- the same upper bound in the Rust provider dispatch, plus its own gate on the identity arm, which returns before the ordinary dispatch and inherits nothing. logos-qt-sdk 6a570b3 (#44) -- the Qt consumer half of the result decode: logosResultFromReply on all three surfaces, so a reply the wrapper cannot read is no longer indistinguishable from a provider that refused the call. The two result-decode changes are paired on purpose and land together here: tightening one consumer surface without the other would make them disagree about one type. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
What
Re-pins module-builder's SDK inputs to the merged full_api fixes:
4775e632f34804anycodegen089142da55fdacanyevent paramsValidation
All five
logos-test-modulesfull_api modules (C++/Rust providers, C++/Rust proxies, ui_qml plugin) build through this module-builder with no SDK overrides.Chain
Unblocks logos-test-modules#24 (re-lock to this) → logos-logoscore-cli#65 + logos-basecamp#264 doctests.
Note: this bumps only the four fixes above. Your socket-stack SDK work (protocol#22, qt-sdk#11) is still open — when it lands, another module-builder bump rebases on top of this.
🤖 Generated with Claude Code