On 10/12/19 12:32 pm, Jeremy Stanley wrote:
(Note that there*are* plenty of high-profile projects on GitHub like Kubernetes using more of a rebase model, with PR owners typically force-pushing changes to address reviewer comments or CI results, and also projects where it's their custom to squash all commits in a PR before merging, so in those cases our normal change counts are more applicable for ~1:1 comparisons.)
Force-pushing is one thing, but projects that squash all PRs in an act of truly evil stupidity (looking at you, CPython) will undercount changes, since what would be a series of changes in OpenStack or a PR with multiple commits in another project ends up as a single commit. That GitHub (and GitLab, who came up with it first) forces you to choose between squashing independent commits in one PR or not squashing the 15 "fix tests" commits in another is a prime example of what makes it an awful tool for code review.