3 ms·
Knative Serverless Kubernetes Is Off to a Bumpy Start
- rektide 5y agokubernetes itself ships releases with a components at a variety of stability levels. i cant see why a project would hold off on 1.0 just because it isnt a total & complete vision & implementation. this article goes extremely hard against 1.0. i can be somewhat sympathetic to "maybe you havent been wrung through hard enough prod uses long enough" but that comaint feels like it disappears quickly & the focus on "how can there be no -stable compnents?!?!" outrage starts fast. this is a disappointingly bot headed read on things to me & insists on idealism, condemning sharply for not finding everything already pristine. as for itcs use- something the article really doesnt touch on- personally event systems feels like a missing serverless/cloud native core. knative eventing seems like one of the only ones out there trying to push forward & make eventing possible within the cloud, within our clouds. there's a lot of "glue" in aws & other clouds that doesnt really exist yet in open source cloud native- queues firing functions hitting endpoints making more events-- knative is a damned fine start towards stitching together the cloud. towards starting to have a way to register things happening, letting other things react. this feels like the basics of the cloud- the thing amid all the services- & knative eventing is a real start at it, a decent one. knative serving i dont know about a whole lot. kubernetes gateway is certainly going to take some of this domain over. but serving is still a real attempt to help, to compose endpoints. the build infra iirc semi-suceeded it's way along fast enough, to where it detached from knative entirely & became it's own reasonably successful project. knative already has offspring. good on it.