Laurent Dumont wrote:
[...] I know that when I merged something for ARA, I ended up appreciating Gerrit but there is a definite learning curve that I don't feel was necessary to the process.
That's a great point. I personally think Gerrit is more efficient, but it's definitely different from the PR-based approach which the largest public code forges use (GitHub, Gitlab). Learning Gerrit is totally worth it if you intend to contribute regularly. It's harder to justify for a drive-by contribution, so you might miss those because the drive-by contributor sometimes just won't go through the hassle. So for a project like ARA which is mostly feature-complete, is not super busy and would at this point most likely attract drive-by contributions fixing corner case bugs and adding corner case use cases, I certainly understand your choice. -- Thierry