korbonits

ack'ed -- I will fix this! -- not the response I am hoping for obviously so I will dial it waaay down.

Thanks for the candid feedback

Which conjectures will be proven false via counterexample next? Dixmier? Poisson?

Proves the need for more formal verification :)

It was Claude Code this January for me. I had heard that it got really good last November but I was still using Kiro at that point.

It wasn't so much an "uh oh" but more of a "eureka!" moment for me. Been busy building ever since.

My reading has dropped off quite a bit in the evenings because sometimes it's more fun to build.

To be honest, I don't know because I have not hit many of those limits due to what I would call "moderate" scale. So far, I have just provisioned enough pods to handle the traffic as-is without using KubeRay. So k8s is handling the load balancing adequately at the moment, but Ray serve is not cluster-aware, only pod aware, for now.

Have you considered using vLLM on top of Ray Serve (on EKS with KubeRay)? KubeRay makes Ray cluster-aware and there could be some optimizations you could make e.g. keeping that GPU fully utilized all the time :)

Sailing.

Low tech, keep the gadgets off (other than VHF 16, and, if it's safe). No sound other than the water hitting the hull. It's quiet like when it snows.

Totally relaxing and rejuvenating. Offline.

You won't want to come back to land.