Hi Dan,
I know the CNCF only has the best
intentions in mind and the survey is meant to better understand our
community. However, there are some inconsistencies that should be
addressed if we are wanting to separate product from project.
As
an example, chart 3 has several things I would consider products
(Oracle, Alibaba), unless OpenShift Online is less than 1% I would see
it as being appropriate to list here.
Also, in terms of storage, unless the cloud native == CNCF, I would expect to see Container Native Storage/GlusterFS (also project not product in the list) in addition to Mino, Rook, etc...
If I may, I
think it might be a worthwhile exercise to review the survey as part of
the TOC to propose a method by which
the statistical representation is more meaningful, as you pointed out.
In addition, clearly defining what we see in terms of a product vs.
project as part of the CNCFs general definitions should be part of this
exercise.
I am sure you can agree, we want to see all projects/products equally represented regardless of the vendor. At this point, given Quinton's sentiment on the shear number or offerings of Kubernetes, it might be in the CNCF's best interest to be more vague in an effort to not unintentionally leave out a product.
Maybe something more along the lines of:
In your Kubernetes environment, do you use:
- Just upstream OSS k8s (DIY)
- Vendor provided k8s platform
- Public cloud k8s service
Where do you run k8s?
- Private Cloud / Data Center
- On public cloud resources
- Using managed k8s service
Thanks,
Erin
