Can I say that I absolutely hate Go's variable declaration! The amount of grief `if err :=` vs `if err =` has caused me is unthinkable.

This is a complicated topic. Microservices have changed the deployment model. But we still use old assumptions to make decisions in this new framework. As such, some of these costs are mostly not understood.

Generate videos in just a few seconds. Try Grok Imagine, free for a limited time.

Or you are running applications where you want deeper insights than CloudWatch is giving you! Esoterics like an open issue on AppMesh to allow custom filters makes me feel that by focusing on the tools, most infrastructure engineers are ignoring very real challenges!.

But shouldn't we focus on the principles of building good infrastructure management systems!. What do you do when your infrastructure scales and performance drops! Or you need to manipulate your traffic beyond what AppMesh/Istio lets you do!.

A significant number of senior infrastructure engineering positions that I look at are focused on the tools. 10-12 years back we were writing C and Python to automate infrastructure. Then we treaded over Ansible, Chef and Puppet to reach the current crazes - Terraform and Pulumi!.

Today I wanted to work on a code base I had been ignoring for a while! I literally had to read through it, to remind myself what the hell I was doing!.

That's a significant change!.

Some alternatives are (I completed this and it is nice!) and

Very useful for understanding programming:

C++ code readability: I find it identifiable. A few years back I spent 6+ months reading a large industrial strength C++ codebase. While it was one of the best engineered software projects I have seen, those were depressing months!.

Tim Roughgarden, ex Stanford and current Colombia, explains blockchains. And one thing he says resonates with me, "Most people's understanding of blockchains are stuck in 2013".

Getting into actually implementing the cluster:.

Finally out with my profile:.

You push one large one through and few smaller ones follow (distributed sequencing):.

Finally first piece of the implementation gets in: Multileader Election.

This is by far the most ambitious project I have undertaken. Linearizability without consensus or quorum!.

Over 3000 people to loose their jobs!!! It stopped being funny!.

I like Musk. He's inspirational even if unpredictable. And I find the chain of events since his takeover of Twitter, while controversial, ticklish.

The tenth part implements an entire TCP server using Tokio.