Hi all,
How do you feel about dropping the contributor and reviewer names from
the commit summary?
Before:
HIVE-28884: Decouple source.q test from SRC dataset (Stamatis
Zampetakis reviewed by Soumyakanti Das, Zsolt Miskolczi, Shohei
Okumiya, Simhadri Govindappa)
After:
HIVE-28884: Decouple source.q test from SRC dataset
The main goal is to increase developers productivity and reduce
boilerplate information.
In many cases the extra information is longer than the commit summary
itself. Every time I merge a PR I have to spend 2-3 minutes editing
the commit message and figuring out the names of every person that is
involved in the PR.
Moreover, the "Author" information is already present in the commit
metadata and the reviewers are clearly shown and tracked under the
respective PR in GitHub so removing them from the commit summary does
not result in loss of information.
The PR id is always present in the commit message (either in the
summary or in the body) so we can easily fetch all the necessary
information (even more and more structured) about contributors and
reviewers of certain PR via the GitHub UI or programmatically via REST
or GraphQL.
For instance the following GitHub GraphQL query can be used to obtain
the name of the author of the PR and the names of the reviewers that
approved the PR.
{
repository(owner: "apache", name: "hive") {
pullRequest(number: 5750) {
title
author {
... on User {
name
}
}
reviews(first: 100, states: APPROVED) {
nodes {
author {
... on User {
name
}
}
}
}
}
}
}
Best,
Stamatis
