inaccurate data for cloud foundry statistics

[Christopher Ferris]

First, it is unclear to me why Cloud Foundry, Kubernetes and Docker have been included as this has primarily been exclusive to OpenStack.

That said, the information underlying the Cloud Foundry statistics is incomplete. While most of the stable code is in the https://github.com/cloudfoundry org, most of the ongoing development takes place in repos under the https://github.com/cloudfoundry-incubator org. Further, in keeping with the same level of coverage as for OpenStack, there is also the https://github.com/cloudfoundry-community org - which is somewhat analogous to StackForge projects for OpenStack.

Further, it is unclear how the underlying data for Cloud Foundry, Docker and Kubernetes are being scrubbed since you don't have access to an analog to launchpad accounts for Cloud Foundry developers that might be able to fill in missing details relating to affiliation. This leaves incorrect representation of contribution.

It would be my suggestion that stackalytics not include Cloud Foundry (or Docker and Kubernetes, for that matter) and remain focused on OpenStack.

[Ilya Shakhat (shakhat)]

Cloud Foundry, Kubernetes and Docker are added to Stackalytics to give OpenStack community an insight in development activity in complementary/competitive projects.

The affiliation of engineers is currently calculated automatically by hostname of email, that's why company distribution data may be not very accurate.

[Christopher Ferris (chris-ferris)]

Then the addition is not serving its purpose if the data are inaccurate.

I don't think it appropriate to include these projects unless the underlying data are accurate reflections of reality; especially given as many if not most engineers use gmail accounts for their github account contact info rather than their corporate email address.