The ironic thing is that Docker became successful only because of the tremendous hype they generated about themselves, and then failed to see the K8s hype train

The ironic thing is that Docker became successful only because of the tremendous hype they generated about themselves, and then failed to see the K8s hype train barreling towards them. Swarm is still what 99% of companies need when they adopt Kubernetes, and is probably 1/10th to 1/20th the cost. Somebody needs to re-invent swarm as an independent project so that we can stop investing in K8s. The cost of K8s isn't even in the complexity, it's in all the millions of side-effects of the complexity, like security, fragmentation, and specialization.