Page
https://insights.linuxfoundation.org/project/argo/contributors?timeRange=past365days&repos=argoproj-argo-cd,argoproj_argo-cd&start=2025-05-18&end=2026-05-18&auth=success
Project
Argo
Issue Area
Contributors
Widget
Contributors leaderboard
Describe the issue
Hi, it looks like the "all activities" contributor stats are unreliable.
I believe this happens mostly because the GitHub "authored a commit" metric is wrong, but I didn't verify the other metrics so potentially there are additional metrics wrongly calculated that contribute to the wrong sum.
Currently, there are cases when users listed in the top 3 had only 3-5 merged PRs to upstream, while contributors with much more merged PRs are ranked lower. My assumption is that maybe commits inside a PR are counted instead of/in addition to merged PRs, but I may be wrong.
Steps to reproduce
You can repro this by choosing Git "Authored a commit" metric and GitHub "Authored a commit" metric individually for argoproj/argo-workflows for the last 90 days, observe that the top contributors are completely different for those 2 metrics, and cross the stats with counting the merged PRs using a GitHub UI query "is:pr author:<ANY_USER_FROM_THE_LFX_TOP_3"
Expected vs. actual behavior
Only merged PRs should count towards the commit contributions.
If there is a metric that counts commits inside a PR, I believe it should be removed.
Jira Issue: DE-954
Page
https://insights.linuxfoundation.org/project/argo/contributors?timeRange=past365days&repos=argoproj-argo-cd,argoproj_argo-cd&start=2025-05-18&end=2026-05-18&auth=success
Project
Argo
Issue Area
Contributors
Widget
Contributors leaderboard
Describe the issue
Hi, it looks like the "all activities" contributor stats are unreliable.
I believe this happens mostly because the GitHub "authored a commit" metric is wrong, but I didn't verify the other metrics so potentially there are additional metrics wrongly calculated that contribute to the wrong sum.
Currently, there are cases when users listed in the top 3 had only 3-5 merged PRs to upstream, while contributors with much more merged PRs are ranked lower. My assumption is that maybe commits inside a PR are counted instead of/in addition to merged PRs, but I may be wrong.
Steps to reproduce
You can repro this by choosing Git "Authored a commit" metric and GitHub "Authored a commit" metric individually for argoproj/argo-workflows for the last 90 days, observe that the top contributors are completely different for those 2 metrics, and cross the stats with counting the merged PRs using a GitHub UI query "is:pr author:<ANY_USER_FROM_THE_LFX_TOP_3"
Expected vs. actual behavior
Only merged PRs should count towards the commit contributions.
If there is a metric that counts commits inside a PR, I believe it should be removed.
Jira Issue: DE-954