-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA512
Hi,
On 1/18/19 6:11 AM, Tyler Cipriani wrote:
In the interim, project-owners are able to opt-in to using the reviewers-by-blame plugin on a per-project basis on their project admin page in Gerrit.
I'm happy to volunteer any of the projects I help maintain as a guinea pig once modifications to the plugin are ready for wider testing.
Also, the Git Reviewer Bot[1] provides folks an opt-in method of volunteering to review a subset of files in a particular repo.
IMO, the main place where the Reviewer-Bot falls short is that it relies on a hardcoded list of file paths that often change and then requires manual intervention. Using blame is nice in that it's automatic .
Getting code review as a new contributor is hard. Thanks for bearing with us.
I've spoken to plenty people that complain that their patch was never merged, and after looking, it had no reviewers assigned :-(
My suggestions for the plugin: * Require some manual interaction by the patch author, similar to GitHub (I believe Paladox is already working on this) * Some account-level opt-out for bots, people who have left the movement, etc. * Don't suggest people that have already been manually removed from the patch * Only suggest people that have been active on Gerrit (or maybe limit it to that repo) in the past X months/years. ** For repositories that have no active development but a maintainer who still checks email, using a time-based check might not work. In a past experiment I looked at the last 100 merged patches. * If feasible, introduce some kind of marking (hashtag? topic?) that can be used in mass cleanup style commits so that they are ignored by the blame so drive by contributors aren't suggested as reviewers. I currently have a mail filter with a few different Gerrit topics (e.g. bump-dev-deps) to separate those review requests from other ones.
Thanks to those that are working on this. - -- Legoktm