docs: update release strategy and command-line flag standardization
- Added a section on the "best effort" release strategy for pre-1.0 versions, emphasizing flexibility and transparency in CI/CD workflows. - Introduced a command-line flag standardization decision, adopting the Abseil Flags library for consistent argument parsing across all binaries. - Outlined a migration plan to phase out the legacy flag system, enhancing maintainability and user experience. Benefits: - Improved clarity on release processes and platform availability. - Streamlined command-line argument handling, reducing technical debt and improving user experience.
This commit is contained in:
@@ -28,6 +28,24 @@
|
||||
|
||||
---
|
||||
|
||||
## Pre-1.0 Release Strategy: Best Effort Releases
|
||||
|
||||
For all versions prior to 1.0.0, yaze follows a **"best effort"** release strategy. This prioritizes getting working builds to users quickly, even if not all platforms build successfully on the first try.
|
||||
|
||||
### Core Principles
|
||||
1. **Release Can Proceed with Failed Platforms**: The `release` CI/CD workflow will create a GitHub Release even if one or more platform-specific build jobs (e.g., Windows, Linux, macOS) fail.
|
||||
2. **Missing Platforms Can Be Added Later**: A failed job for a specific platform can be re-run from the GitHub Actions UI. If it succeeds, the binary artifact will be **automatically added to the existing GitHub Release**.
|
||||
3. **Transparency is Key**: The release notes will automatically generate a "Platform Availability" report, clearly indicating which platforms succeeded (✅) and which failed (❌), so users know the current status.
|
||||
|
||||
### How It Works in Practice
|
||||
- The `build-and-package` jobs in the `release.yml` workflow have `continue-on-error: true`.
|
||||
- The final `create-github-release` job has `if: always()` and uses `softprops/action-gh-release@v2`, which intelligently updates an existing release if the tag already exists.
|
||||
- If a platform build fails, a developer can investigate the issue and simply re-run the failed job. Upon success, the new binary is uploaded and attached to the release that was already created.
|
||||
|
||||
This strategy provides flexibility and avoids blocking a release for all users due to a transient issue on a single platform. Once the project reaches v1.0.0, this policy will be retired in favor of a stricter approach where all platforms must pass for a release to proceed.
|
||||
|
||||
---
|
||||
|
||||
## 📚 Full Workflow Reference (Future/Formal)
|
||||
|
||||
The sections below document the **formal Git Flow model** that yaze will adopt post-1.0 or when the team grows. For now, treat this as aspirational best practices.
|
||||
|
||||
Reference in New Issue
Block a user