Ease access to CFF contribution statistics

[gberche-orange]

As a CF user

• in order to assess community health
• in order to use data to understand vendor involvement contribution
• I need to easily have access to statististics and data about community contributions (in particular code, and handling issues, reviews of PRs)

As an inspiration, the CNCF hosts the https://devstats.cncf.io/ for CNCF projects

Full list of dashboards

Community sizing and health assessment dashboards * Companies contributing in repository groups * Companies table * Company Statistics by Repository Group * Countries stats * Github Stats by Repository * Github Stats by Repository Group * Overall Project Statistics * Stars and Forks by Repository Contributor statistics dashboards * Bot commands repository groups * Company PRs in repository groups * Contributions chart * Developer Activity Counts by Companies * Developer Activity Counts by Repository Group * New and Episodic PR Contributors * New contributors * PR Reviews by Contributor * PRs authors repository groups * SIG mentions Issue velocity dashboards * Inactive Issues by SIG * Issues age by SIG and repository groups * Issues Opened/Closed by SIG * New And Episodic Issue Creators PR velocity dashboards * Awaiting PRs by SIG * Blocked PRs repository groups * Inactive PRs by SIG * Open issues/PRs by milestone and repository * Open PR Age by Repository Group * PR Comments * PR Time to Approve and Merge * PR Time to Engagment * PR Workload per SIG Chart * PR Workload per SIG Table * PRs approval repository groups * PRs labels by SIG * PRs labels repository groups * PRs opened by SIG * Other * Licenses and programming languages * Repository groups

Sample screenshots

with [Community sizing and health assessmentCompany Statistics by Repository group](https://k8s.devstats.cncf.io/d/8/company-statistics-by-repository-group?orgId=1&var-period=q&var-metric=contributions&var-repogroup_name=All&var-companies=All)  per project statistics, e.g. [cluster api](https://k8s.devstats.cncf.io/d/24/overall-project-statistics?orgId=1&var-period_name=Last%20year&var-repogroup_name=SIG%20Cluster%20Lifecycle%20(Cluster%20API)) 

Fortunately, the code powering devstats.cncf.io is opensource and could be used as is by the CFF.

See related KubeCon session Who What How: Understanding Kubernetes Development through DevStats with video and slides

Associated code is at https://github.com/cncf/devstats and https://github.com/cncf/devstatscode

[chipchilders]

Is this useful? https://cloudfoundry.biterg.io/app/kibana#/dashboard/Overview

Also take a look at the LFX Insights page for CF (DATA NOT CONSIDERED READY): https://insights.lfx.linuxfoundation.org/projects/cloud-foundry%2Fcloud-foundry/dashboard

One of our challenges is that the paired programming model necessitates sometimes "splitting" commit attribution amongst two or more people. That was implemented for the CFF by Bitergia a few years ago. As the project progresses, this may or may not remain an issue... but we are assuming that pairing will continue (and is obviously of historical significance). We worked with the LFX Insights team on this feature over the last year, and are starting to validate the data that they are pulling in. That work is not done yet.

We can support whatever the technical community would prefer long term, so this is a great discussion to have!

[gberche-orange]

Thanks @chipchilders, this dataset is pretty useful!

Was it shared with the community a long time ago (sorry if I had missed the announcement)?

The dataset seemed accurate to me regarding my company (Orange) including my own contributions.

Some feedback of potential improvements (besides paired commit issue you mentionned) covering use-cases I did manage to cover in the tool:

• support qualitative analysis: enable drilldown to actual data in dashboards: e.g. PR/issue/email URL/Subject
• monitor responsiveness to community contributions (PR and issues),
  • PR Time to Engagment
  • PR Time to Approve and Merge
  • Issues without comments/responses (excluding automated response such as cla bot and pivotal tracker bots)
  • Time to Issue comments/responses (excluding automated response such as cla bot and pivotal tracker bots)

[chipchilders]

@gberche-orange - Thanks for the feedback!

The bitergia dashboard was put in place back in 2015, so that was a long time ago. :) We don't highlight it much, because vanity statistics are always an apples to oranges comparison. We also deemphasized it when the community started getting flooded with very low value PRs in an attempt to game the system.

The LFX Insights page / dashboard will get more promotion once we are comfortable with the data and features there, including adding it on the main cloudfoundry.org website. It's still early, so no... no community "announcement" about it yet. I'll make one when I'm satisfied enough with the quality (sometime early 2021) and when we have incorporated it into cloudfoundry.org.

For your specific needs, here are some Bitergia pages that might help... we will work with the LFX team to have similar stats. FWIW, CNCF is also working with LFX to implement many of the most commonly used features of devstats.

• Drilldown - You can drill into the PR's in the dashboards below
• PR Responsiveness
  • Backlog
  • Timing
  • Efficiency

[chipchilders]

Hopefully I answered all of your questions @gberche-orange ... I'm going to close this issue now, but let me know if there is something specific that's outstanding with it!