Release year: 2023

Author: Gergely Orosz

Buy this book (note: affiliate link)

Link to my handwritten notes

Review

I envy people who choose a profession like software engineering, because it feels like the career path of such professions is so clear: from junior to senior engineer, to staff, to manager, director, VP of Engineering, to finally CTO. In a sense, this career ladder feels like playing an MMORPG, which is supposed to be quite rewarding and addicting as you start your climb.

The problem is that some part of me refuses to fit this mold, and I can’t put my finger on why. This is why I chose to read The Software Engineer’s Guidebook by Gergely Orosz, an Amazon bestseller, to reframe my mind on software engineering. Maybe this book could become my cure to start better understanding how software engineers should work and think about their work and, through better collaboration with them, I could start figuring out my own path.

If you’re interested in getting this book, here’s what you can look forward to:

  • a broad overview of the expectations that come with being a software engineer
  • tips for being a great contributor
  • what mindsets give the best results
  • how to grow from an individual contributor into a manager

The book certainly has merit, being written from an experienced engineer who spent a lot of time working at companies like Uber and who spent many years writing on the topic on his blog . However, don’t expect depth; its length is used to cover a multitude of topics connected to software engineering, not to be the ultimate resource for each topic. In a sense, the book is an introduction to all the topics it discusses, and expects the reader to search for more details if their interest is piqued. For example, on topics like CI/CD, the author will explain what the concept is and why it is useful, but will not guide you into building your first pipeline (though he often provides useful links for further reading).

I think this is a solid book to have in any organization’s library, because surely it will have some software engineers among its wall and all of the topics discussed in this book are important. Now it’s only a matter of diving deeper in each of those subjects to correctly put them into practice.

Félix rating:
👍

⭐ Star quotes

  1. (p. 40) Good feedback is:
    • timely
    • specific
    • actionable
  2. (p. 55) Keep track of your achievements. The most practical way to do this is to write down what you do, every week. Summarize key achievements each month and quarter.
  3. (p. 67) Done is better than perfect, especially during wartime.
  4. (p. 94) The easiest way to turn down a request is to reply “yes, I’d like to help, BUT…”
  5. (p. 157) Create a system for capturing important but not urgent work. This could be a to-do list, a simple document, a physical notebook, …
  6. (p. 164) Once you have a full understanding of how things work, “thinking outside the box” will simply be offering the solution that experts tend to do.
  7. (p. 174) “Every mentor I worked with ended our conversations with a ‘you got this!’ or ‘you can do this!’ or similar. Only words, but damn, they meant the world to me.”
  8. (p. 176) Asking honest questions before giving feedback gives the receiver the opportunity to share more context, with the absence of a negative sentiment.
  9. (p. 225) Kickoffs are events where big questions are clarified with relevant stakeholders.
  10. (p. 229) The Physics of software projects: In the trio of “Timeline, Scope, and People”, if one of these things changes, at least one other also needs to change in response.
  11. (p. 270) Processes are never the goal. Don’t ask which processes your team should have in place, ask how the team can get stuff done better and faster.
  12. (p. 273) One easy way to know if there is clarity is to ask engineers what the team’s goals are, and why these goals exist. If everyone gives roughly the same answer, there’s clarity.
  13. (p. 281) As a tech lead, not every problem is yours to solve. For example, a colleague with performance issues isn’t your problem; that’s on your manager.
  14. (p. 287) A good resource for mapping our which career levels companies have beyond the senior engineer level is the website https://levels.fyi
  15. (p. 289) With any collaboration, the key to making it work is trust.
  16. (p. 296) “Why do customers use us?” is an obvious question, and it’s surprising how often engineers don’t have an answer.
  17. (p. 311) $$\text{Trust} = \frac{\text{credibility} + \text{reliability} + \text{authenticity}}{\text{perception of self-interest}}$$
  18. (p. 323) When not deeply embedded in a team, aim to adapt to it rather than reshape it around you. Colleagues will respect you more if you adapt and help them improve practices.
  19. (p. 332) A service catalog is a portal where teams can register their services, and engineers can search for them. It should answer these questions:
    • Is there a service that does X?
    • Who owns service Y and where is the oncall rotation?
    • How do I onboard my team to this service?
  20. (p. 342) Logs are meant to help an engineering team debug production issues, by capturing missing but necessary information for future reference during troubleshooting.
  21. (p. 346) Business metrics:
    • Customer onboarding
    • Success and error rates for business-specific operations
    • Daily/weekly/monthly active users (DAU, WAU, MAU)
    • Revenue
    • Usage numbers: how long does a user interact with the app?
    • Number of support tickets
    • Retention and churn
  22. (p. 347) A noisy alert is one that’s not actionable.
  23. (p. 347) Precision measures the percentage of alerts which indicate a real issue. A system with 30% precision means 3 in 10 alerts are outages, and the rest are noise.
  24. (p. 348) Recall measures the percentage of outages for which alerts are fired. A system with 30% recall means alerts are fired for 3 in 10 outages.
  25. (p. 348) The ideal oncall system has 100% precision with no noisy alerts, and detects 100% of outages.
  26. (p. 348) ⭐ Have the oncall engineer record whether each alerts is for an outage, or if it’s noise.
  27. (p. 348) ⭐ In incident reviews, go through all recent outages and answer the question: “did an alert fire which indicated an outage was happening?” This will show the recall percentage.
  28. (p. 355) A common mistake less experienced engineers make is trying to understand why an outage occurred, and to only start fixing it when they know. If there are obvious mitigation steps that can be started, like rolling back a code change, or executing a rollback, do these first!
  29. (p. 358) Consider alerting on unknown states. A state that’s neither good nor bad tends to be a hotbed for problems later.
  30. (p. 360) When it comes to architecture, you should use formal approaches only when you’re convinced it adds value, and team members understand it.
  31. (p. 382) Always understand why you are working on something, and how things works:
    • Why are we doing this project, who benefits?
    • Why does it work now: what changed?
    • Which alternatives could we use?
  32. (p. 386) If you really want to force yourself to learn something well, give a presentation, or hold a session about it.