I often agree with Kelsey Hightower, but there are so many things he doesn't mention here. For example, being able to independently deploy components frees up c

I often agree with Kelsey Hightower, but there are so many things he doesn't mention here. For example, being able to independently deploy components frees up certain kinds of development worklows. Distributed components also scale and fail independently, and you can use nifty things like message queues between them to provide resilience and soak up load spikes. I'm sure the pattern has often been applied in the wrong use cases, and that many people have over-applied it, but "the monolith is the future" seems just as wrong as "microservices are the future." We are nowhere near the size of a large bank... or even a small bank, and yet we've benefited from a distributed set of independently deployable and scalable components. You can call them microservices, or not. I can think of ways we could restructure on a monolithic backend, but just noodling on the idea leaves me with more constraints than benefits. Idk, it's a thought-provoking statement at least, but I sort of wish we'd stop reacting to fads with anti-fads.