Exactly what I'm always saying. It's also nearly impossible in an Enterprise IT environment to get Kubernetes working on your laptop. Minikube and Docker Edge both seem to fail way too often.

As a developer one wants to spin up a system to work on, then work on it, then push results to some repo. And this loop simply isn't possible (yet?).

Also what the author didn't mention is that even the vanilla k8s stuff is already super complex. Let's assume you manage to setup a cluster somewhere and it continues to work for more than 2 days (rarely seen in real world work environments). Then you are faced with deploying your hello world app to work on. Just for a simple single-server nginx deployment with no files and no config, you already need to understand mutliple objects: deployments, replicasets, pods, containers, nodes, hosts, services, nodePorts, port-forwards, maybe even ingresses and controllers.

That means until you feel well, even in a perfect environment, you need several days or weeks. And the documentation is not really helping you there. Yes, it's better than most enterprise grade documentation out there, but still it assumes a lot of stuff upfront. For instance, why should a developer even know what an ingress is and that it might be something he needs?

Combine that huge learning overhead with nearly impossible network debugging, beta level stability, and the low possibility to "just use it", then you have a system that most developers will never touch.

Docker itself might survive though, and one of the good things of this CNCF world is that other alternatives to docker also get a chance to improve on the existing system.