Hacker News Reader

Comment by aaronbrethorst | Hacker News Reader

I stopped buying technology books years ago because—too often—they focused on the latest and greatest ("covers the latest .NET Framework, updated for v1.1!"), and therefore are rendered obsolete within months.

Write about the "why's" instead. In other words, try to answer the big picture questions about serverless architectures without getting bogged down in the minutiae of what is state of the art in early 2016. Try to picture what someone in the year 2018 or 2020 would want to know about them, and write about that.

View on HN

This. Trends change so fast that even if you write the best book of it's kind it may be irrelevant in the next year. Talk about why this paradigm is fundamentally different. Now that developers have to worry less about server provisioning, capacity constraints etc. why is that important? What flexibility and capabilities do we have now that we didn't before?

Agreed.

Great points.

We have dedicated many pages (and will dedicate more) explaining how/why serverless is different. Why we think it's superior (in some instances) to traditional technologies and how it can alleviate the burden that comes with managing traditional infrastructure (and current methods of implementing back ends for web applications).

The book is covering AWS and technologies such as Lambda, and API Gateway. These have already evolved as we began to write the book. I want to focus more on patterns and applications but we also need to introduce readers to these patterns and show how to actually put them together - so it's an interesting question of finding balance.

Don't just tell people why it is different. Tell people why it is the same. What is timeless about it? How do I implement the Gang of Four's singleton pattern? [well maybe that's not a good place to start].

How about composability as an organizing theme?

Good luck.

As the parent has said, please focus on high-level principles and avoid going deep into the current state of the art.

I have found that the only transferable knowledge I have over years of programming, are the language-agnostic high-level / 'core' principles

Reply on news.ycombinator.com