List view
Before we launch the clients we want to make sure we have not missed large features or subtle corner cases in the generated code. Serialization / deserialization could have unknown bugs. Building HTTP requests and responses with different data types could have problems too. And finally we may have bugs that do not manifest until one runs for many hours or days. We need to test all these things.
No due date•26/26 issues closedWe also want to stabilize things exposed by the generated APIs, like `gax::*`, `auth::*`, and `wkt::*`.
No due date•74/74 issues closedWe want to get the secretmanager crate to a launchable state. We may or may not launch it, we just want to remove any technical blockers from doing so. * We want the integration tests to pass * We want the documentation tests to pass * We want an example showing how to initialize the client * We want a README.md that includes this example * We want the basic structure of the client to be stable, e.g., traits for the client, builder, and/or options * We want `sidekick refreshall` to work * We want `sidekick upgrade` to work * It is not a goal to have an automatically generated quickstart * It is not a goal to have a stable API ready for 1.0 * It is not a goal to automatically detect breaking changes * It is not a goal to automatically update the code as googleapis changes
Due by March 1, 2025•22/22 issues closedThe generator can generate a working secret manager client library for Rust. For Golang we want compilable code. By "working" we mean that the client libraries compile without warnings and can issue RPCs with the default ADC. We do not require perfectly clean "lints". We do not require working reference docs. We do not require packages. We do not require code samples. We do not require retry loops. Secret manager does not include LROs. We do not require telemetry headers, user-agent strings, or other nice-to-haves.
No due date•19/19 issues closed