Distributed systems programming has stalled

I agreee it has stalled, I think for almost everyone what the author considers bandaid is practical enough that isn't a need for innovation. Distributed systems is more or less solved imo.

Agree with another commenter, observability tools do suck. I think that's true in general for software beyond a certain amount of complexity. Storing large amounts of data for observability is expensive.

I'm very impressed with the quality of some observability tools like Datadog, which do many good things either automatically or very easily. The usability is leaps and bounds ahead of things like New Relic or the manual instrumentation intensive open source tools. But yes, the costs are insane and require some diligence to keep from running too wild, like most SaaS products these days.

But ultimately we pay it because it gives us incredibly valuable insights and has saved us countless hours in incident response, debugging, and performance profiling. It's lowered my stress level significantly.