Thread by @QuinnyPig

So this caught fire, so let me go into some depth here.

I'm not wed to the timeline--I am wed to the outcome. A thread...

https://twitter.com/QuinnyPig/status/1093261169614356490

I was once very militant about the OS, and later, the distro. Then I was militant about the web server you ran on top of it.

Today none of that is even slightly relevant.

"But wait!" you interject. "It's super important to what I work on because..." and that's where generally we diverge.

It may well be relevant to you, but it's not relevant to business value.

Today there's no meaningful difference to a business what the infrastructure really looks like; what's valuable is what that infrastructure does.

Your company exists to solve a problem. If that problem isn't "infrastructure," then the bits are the "how," not the "what."

The reason I'm bearish on Kubernetes in the long term comes in two parts.

The first is that it too will slip beneath the waves of "plumbing I don't have to care about." That's the easy prediction.

The second, and much likelier to result in hate mail, is that I've yet to have a conversation that answers the question "what is the business value of Kubernetes?"

"It lets us do X and Y and Z more effectively"--okay, great.

Kubernetes speeds time to market. That's hugely valuable, except I don't see it happening in the world.

I see companies getting bogged down in the weeds. I see companies hiring consultancies, expensive staff, or providers all to handle the sticky bits for them.

If that's you, I have a question. Has that investment bought you the velocity you'd hoped for?