Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

I really don't like either idea, and don't understand why is this such a big issue...

Working with Web interface for anything code-related is a huge downside. Workflow doesn't matter at this point, the usability of this approach, no matter what it does is so bad, it's not worth discussing further.

The best way it ever worked for me is that:

1. PR author creates a PR and gets assigned a reviewer.

2. The reviewer leaves a commit with

    # REVIEW(reviewer): Comments
3. Then the PR author changes something or argues back.

4. If reviewer is happy, the author gets to organize the commits in whatever sequence they want (this would typically involve something like squashing everything, removing review comments if they are no longer necessary, and then splitting the PR into logical parts). Otherwise we go back to (2).

No need for complicated workflows, no need for any kind of external system...



Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: