I bought the book with high hopes, but to be honest I found it a little dry. About halfway through I had to give up on the idea of reading it cover-to-cover, and instead just cherry-picked the interviews I wanted to read. It lacks a lot of background info on the interview subjects, so unless you are familiar with them you may find yourself lost as many historical and technical details are referenced without additional explanation.

Personally, I thought "Founders at Work" was a much more interesting read and would recommend checking that one out first.

I read Coders at Work and enjoyed it. As someone who picked up programming in a post-web age it was very interesting to read about our heritage. I bought the book expecting nuggets of architectural and project wisdom but found less than I had anticipated in that respect.

Yes, reading specific interviews is one way to read it.

Another way is reading every interview and comparing answers to the common questions Peter asked. Some answers are close to home, some unique, all interesting. Comparing them may be educational to all.

I'm busy reading both now and I'm in total agreement. I am, however, reading it cover to cover.

I've learnt a lot, but that's because I went and read up about the topics that I didn't know about.

I'm also enjoying founders at work more.

Yeah I have found that, stumbled across quiet a few things I knew nothing about, looked them up and learned even more in the process.

It depends on your experience I guess. I'm reading it now, and up to Joe Armstrong's interview everything is very understandable.

I was a bit disappointed the author didn't really ask Joshua Bloch about the non-Java JVM languages (in his answers Bloch thinks about Java and the JVM as a whole, while the discussion on concurrency for example would clearly benefit from separating the two).