No way to incrementally improve linters
- 主要语言
- Python
- 星标
- 50
- 派生
- 129
- 平均合并
- 3 天 10 小时
- 30 天内合并 PR
- 3
描述
As discovered when doing https://github.com/ament/ament_lint/pull/174, the expectation of making linter changes is that, even when fixing linters to comply with documented coding standards, there is an expectation of wide-ranging PRs in parallel to accommodate these changes.
It seems we need a better path forward for linters already in use. Some discussion here: https://github.com/ros2/rclcpp/pull/1052
Here are some suggested fixes:
(1) allow specifying a reference commit so style issues before that commit are grandfathered in, preventing code churn.
(2) version the coding standards so that packages have to affirmatively opt in to later refinements, and they don't automatically create CI regressions.
贡献指南
评估
这个 Issue 还没有评估数据。