Thoughts on the book "Accelerate" | ✉️ #19

Alt: Illustration for MKDEV Dispatch #19 featuring a smiling man holding a small dog, with text "Thoughts on the book 'Accelerate'" on an orange and purple background with paper plane graphics. Alt: Illustration for MKDEV Dispatch #19 featuring a smiling man holding a small dog, with text "Thoughts on the book 'Accelerate'" on an orange and purple background with paper plane graphics.

Hey! 👋

I rarely read IT Books. As someone who spends anywhere between 40 and over 60 hours a week immersed in IT-related work (which also includes a lot of reading - of documentation, of code, of articles), on my spare time, I prefer to read about something else. Still, now and then appears a book that everyone is talking about and that seems to become an industry must-read. I usually wait for a couple of more years and, if it’s still being recommended to me, then I take my time to sit and read such a book.

This month, I finally read “Accelerate”, a book about what distinguishes high-performing from low-performing organizations, and what enables them to move at a much higher speed.

At first, I couldn’t stop thinking that everything the book says is kind of already known and obvious to me. It talks about how certain practices, like source code management, or agile, or CI, are more frequent at high-performing organizations. All of this, and more, we at mkdev bring to our customers on a daily basis. Moreover, “Accelerate” doesn’t give any deep explanations, or examples, or practical advises. It just states “X is good”, and won’t tell you how to get to X.

But then another thought hit me: the fact of popularity of this book alone shows that all the things, that I take as obvious truths, are far from being them. Majority of the companies and projects are still not there yet, there are still many environments that can’t move fast and that didn’t embrace the new approaches yet. And the first step on this path is to understand what needs to be changed and why. “Accelerate” is good at doing both, so I will probably recommend it in scenarios, where the theory of innovating fast and without compromises is not yet understood.

What We've Shared

  • Is Bard better than GPT-4? Few weeks ago we saw Sundar Pichai explaining how good is bard, and how fantastic is PALM2, but is that true? Are we sure about that? Today we are going to explain what we've seen with the information that we have.

  • GitHub Copilot X vs GPT-4 for DevOps work - are they really the same? GitHub Copilot X is the next version of coding assistant from GitHub, which promises GPT-4 and nicer developer experience. But is it really as good as ChatGPT with GPT-4? In this video, we are going to try to do the same DevOps task with both and compare the results.

  • Should I use a Cloud Provider? Cloud providers' collective revenue exceeds $200B, reflecting their vital role in our digital world. Pablo Inigo Sanchez delves into the shift towards cloud computing, considering its pros and cons across various sectors. Most find its scalability, cost-effectiveness, and shared responsibility appealing, making it a compelling choice for many businesses.

  • Why Do You Even Need Helm? Kickstart your Helm journey with our Helm Lightning Course. Grasp Helm's fundamentals, its role in Kubernetes, and its deployment capabilities. From handling YAML files to the importance of Helm charts, we cover it all. Ideal for those looking to optimize their Kubernetes experience.

  • 'DevOps Accents', episode 11: Cyber Security, Open Source and Cloud Costs.

What We've Discovered

  • CloudMapper: One more - and relatively old - tool to analyze your AWS environment, including building a diagram of it.

  • How Canva saves millions annually in Amazon S3 costs: An excellent walkthrough of S3 cost optimizations, which essentially boils down to "knowing how your data is being used".

  • Aurora I/O Optimized Tier: As much as we love Aurora, the I/O pricing of it can bite you pretty hard. New tier removes this in exchange for a higher per-hour price for instances themselves, which can easily shave 10-20% off your monthly bill (for certain workloads, of course).

  • Kubecon Amsterdam Learnings: A sum up of recent Kubecon. Essential reading for understanding the latest trends and developments in both Kubernetes and DevOps worlds.

  • The Principles of Designing Microservices: A good small intro on how to do microservices properly.

A random reminder

All previous episodes of our podcast are available on its dedicated page. You can now read the descriptions of every episode right on the website. Choose what you want, or go through all of them!

The 20th mkdev dispatch will arrive on Friday, June 9th. See you next time!