Hacker News new | past | comments | ask | show | jobs | submit login
Ask HN: Books you should have read when you start a career in SE / CS?
216 points by DDerTyp on May 9, 2018 | hide | past | favorite | 95 comments

Three boks I'll recommend you read as soon as possible.

1) The Mythical Man month

3) The Pragmatic Programmer

3) Code Complete.

The thing about these books, well atleast 1 and 3 is that when they were written they broke new ground. Or if that phrase bother you, they were the first mainstream publications that brought their core ideas to developers.

Reading these books early in your career will help cement these best practices.

Reading these books after you've been a practicing programmer means,IMHO, that the return you get from reading them goes way way down.

You'll find that you've already learned all/most? of what they are trying to teach.

You'll find yourself getting frustrated that they are preaching what you already know as common sense and first principles.

Or put another way, I'm not sure there is much that someone who has programmed for 5-10 years can get out of those books that they haven't picked up somewhere else already or learned the hard way.

A quick reference for the important points in Pragmatic Programmer :https://blog.codinghorror.com/a-pragmatic-quick-reference/

Code Complete: http://codecourse.sourceforge.net/materials/Code-Complete-A-...

Thank you. The most value I get from these kinds of books is to become aware of important concepts so that I can actively work on improving them. In other words: what I seek from these books is the mental framework that experienced developers rely on, the rest is mostly noise.

I read through Code Complete 2 cover to cover my first year and I agree with parent it's more useful early on than later. There's a lot of discussion around best practices; I think there's an entire chapter on strategic whitespace. All the stuff in there makes sense, but it cements why it's important.

Definitely check out "The Pragmatic Programmer"! There's definitely some good nuggets in there that helped me. Distinguishing full-blown prototypes vs. prototypical kernels that can be productionized, a.k.a. tracer bullet prototyping, is one.

This x1000. I'm only just getting to reading these now 10 years into my career and I find them to be a bit useless for me since I've already picked up bits and pieces of each throughout my past jobs/companies. Would have been much better off earlier on if I had just read these beforehand.

I kinda wonder, after all these years, if there would ever be a new edition of the Pragmatic Programmer; or even a discussion from the authors on how their thinking has evolved over time.

I always started and stopped reading Code complete book. The most recent strategy that really worked for me was. I googled summary notes for that book to recognize the topics that I wanted to learn and ditched the ones that are redundant.

This rejuvinated my interest and productivity for reading that book.

Some I haven't seen mentioned here:

1. Coders at Work

Peter Seibel does an amazing job of asking programmers questions that make them explain their methodology. The interview with Donald Knuth is awesome; really enjoyed hearing him talk about literate programming.

2. The Soul of a New Machine

Tracy Kidder's 1981 Pulitzer Prize winner I think is a brilliant case study on how engineers work together and the things that can go wrong and right with different personalities interacting with highly technical ideas. The project in this book starts without the consent of management, which to me shows the value of questioning the system to add business / engineering value.

3. Little Schemer

Small book that will give your brain a serious workout and show you how to problem solve with Lisp like languages. Even if you never end up using a Lisp dialect, this book expands your brain.

I have more here, all of which I recommend for software and other careers http://benbrostoff.github.io/books/

+1 for Little Schemer. Before I read it, I could write recursive methods, but after I read it I could think about recursion naturally.

This! The primary benefit of this book is in learning to understand recursion. Lisp just happens to be a good environment for naturally expressing recursion.

+1 to Coders at Work. That book does an excellent job of showing the human side of programming.

I think too many people approach the industry thinking programmers are god-like deities who have an unlimited amount of time and brain power. Coders at Work does a great job breaking down the "imposter syndrome" plaguing new people coming into the industry.

The Soul of a New Machine is a wonderful book. I don't think it helped me career-wise, but it's an absolute joy to read.

Don't Make Me Think by Steve Krug and The Design of Everyday Things by Don Norman.

While they're not strictly software or programming related, I really like the concepts of making interfaces that require minimal thought to use and empathizing with your users. I've seen an unfortunate amount of "programming machismo" where a confusing or poorly engineered system is used and accepted because "that's just how things work in the real world". And whenever someone struggles with the process, it's obviously because they're wrong, not the system. But more often than not, if more than one person has trouble understanding your system, whether it's an API, a website, a build process, the design is the issue.

Don't Make Me Think is a good one, but I always felt like its lessons could have been summed up in an infographic...though I guess that would be missing the context to get people over that hill. I guess if you don't know that intuitively, you're the kind of person that needs that book.

I honestly felt the same way about "The design of everyday things". The first few chapters are quite interesting but it gets insanely repetitive after a while. To be fair I feel that way about a lot of pop-sci books...

Our field is so varied that there is no good answer to this question unless you specify what you want to focus on.

Some people are inevitably going to recommend The Art of Computer Programming, which hardly anyone has read and isn't that relevant to the work that 99% of us are doing.

Someone will probably recommend The C Programming Language, also called K&R after the authors, but again it's not very useful unless you're going to be using a lot of C and even then I personally don't think it provides anything you can't get from guides on the internet. (I have actually read this one, but it's been a while)

There are other classic textbooks that will probably be mentioned that are only useful to those in that domain. Many of us can get by without a deep understanding of algorithms and data structures, most people don't need to read a compiler book, etc, etc.

However, there are some generalized books that people often recommend like The Pragmatic Programmer and Code Complete. These are, in my opinion, good recommendations if you're looking for something to read but I wouldn't say that someone starting a career in software engineering should read them. Software engineering isn't really a field with seminal texts that should be read by everyone.

For example, every political scientist should have probably read The Republic by Plato and Politics by Aristotle among many other texts. I don't think software engineering has that equivalent, partially due to the relative newness of the field and partially because the primary output of our field is not written texts.

Another interesting question would be "What codebases should someone starting their career in software engineering look at?". Are there things that the average joe programmer can learn by poking around the Linux kernel or Firefox? Maybe. I don't know. I've never done it. I'd be interested in hearing from those who have, though. Maybe I'll do my own ask HN.

edit: I feel I should clarify that this is written from the perspective of a software engineer, not an academic in computer science. For all I know there could be seminal texts that those on the academic side should all have read.

"The C Programming Language" is a good read if you intend to write a book. It is well written so you should read it for inspiration on how to make your book good.

If you want to learn C - there are plenty of resources that work just as well. The good ones cover things that didn't exist back then. (I've only read the original K&R from before ANSI, not one of the latter edition(s) so I can't comment on how up to date the latest version is)

> recommend The Art of Computer Programming

And, if you do decide to dive in and read them, plan to spend a few years on them.

It works as a reference, just today I found out that the devops darling of anomoly detection [1] is in volume two [2].

[1] https://www.slideshare.net/tboubez/5-things-i-learned-toufic... [2] https://en.wikipedia.org/wiki/Kolmogorov%E2%80%93Smirnov_tes...

I sort of wish my CS grad school had just been studying those books versus some of the stuff I did cover.

Designing Data-Intensive Applications[0] by Martin Kleppmann. There's a previous HN thread about it[1]. Helped me understand a bit more about databases and systems. The book is also very approachable and has the perfect blend of application and theory at a high level that anyone approaching the industry for the first time stands to gain a lot from reading it.

The Architecture of Open Source Applications[2] series is a good one for leaning how to build production applications and you can read it online. The chapter on Scalable Web Architecture[3] is a must-read.

[0] https://www.amazon.com/Designing-Data-Intensive-Applications...

[1] https://news.ycombinator.com/item?id=15428526

[2] http://aosabook.org/en/index.html

[3] http://aosabook.org/en/distsys.html

As far as a technical book, Data Intensive Applications by Martin Kleppman. This book covers distributed systems as well as databases, data-warehousing, and data-processing.

A lot of tools and patterns you don't see in school that describe how most of the "magic" behind high-powered software we use is implemented.

This should honestly be the top-voted answer. The standard palette of cargo cult options are at best a weak proxy for real-world experience.

But Kleppman's book is genuinely and outstandingly useful for understanding the various components involved in modern backend architectures.

My take on some which I think should be atleast read once:

  1) The unix programming environment
  2) The design of the unix operating system
  3) The Pragmatic Programmer
  4) Programming Pearls
  5) Computer Systems: A Programmers Prespective
  6) K & R C
  7) The art of Unix Programming.
PS: I admit my bias towards systems.

"The unix programming environment"

I read this in like a week at one of my first jobs. I learned so much that in a few days my status in the team radically changed and became the go to person when people's scripts didn't work; word spread and soon I had people from other teams seeking help with their scripts too. No one else read the book though AFIK ...

What I really liked about this book at the time:

- it explains the few key concepts extremelly well

- the exercises were truly outstanding, really testing your understanding of subtleties of above concepts

One for the ages.

I like (1) The unix programming environment

It's very old and underrated. Many people have problems getting started with the command line and basic UNIX tools. This is _THE_ place to learn about this.

I’m halfway through the 5 and its an amazing resource.

Really phenomenal textbook. Maybe the only one I've read (almost) all of the way through.

I've read most of the recommended books, but as far as what I think has delivered immediate real value, I'm going to go with:

1) Clean Code 2) The Phoenix Project 3) UNIX and Linux System Administration Handbook

I'm sure #2 will be something folks want to fight over shrug. If you want to sidestep years of headache it's a must-read though.

FWIW, "The Goal: A Process of Ongoing Improvement" by Eli Goldratt, almost certainly was the inspiration for The Phoenix Project, and is IMHO a better read. One could follow up with any of Goldratt's other books for more information about the Theory of Constraints.

In a sense, I agree with you, but TPP is specific to a software organization and I find it "translates" better, especially to the inexperienced.

You make a compelling argument that making the information accessible is more important than the means of conveyance.

This is a good point for many books listed in this thread: it doesn't matter how you pick up the design pattern knowledge, the algorithm knowledge, the code quality knowledge, etc etc, so long as you get it, and with complete understanding.

Couldn't agree more. The fact that there are options out there means that there are more ways for these things to hopefully "click".

Another vote for The Phoenix Project. It really helps to set the right mindset for identifying and addressing the multitude of non-technical challenges that quickly become the focus of many (I'd venture most) developers' careers.

To elaborate, the best things you can do for yourself when starting your career are be easy to work with and have a clear picture of what good code/system/project/team/company organization looks like.

Don't worry about coming out of the gate skills-hot. It's a marathon, not a sprint. If you are lucky and good, you will be doing your best work late in your career, not at the beginning.

None of the other recommended books tell you how to name things, and that's always going to be the most important problem to solve anyway ;)

Structure and Interpretation of Computer Programs:


This is the most influential book I have read in my SE career. It's not a LISP book. It's a book about programming fundamentals. E.g. Objects are covered as well. They even build a register machine at the end.

I think a translation to a more modern language (lua!) would boost adoption quite a bit.

How To Design Programs pursues a similar goal. Esp. 2nd edition has largely been rewritten and uses a more modern programming language. Maybe give it a try?

There are many books which sound good in theory but I think my younger self would not have been able to make use of the wisdom in those books because I did not have the experience to relate to what the book was talking about. Nevertheless, if I had to ignore this factor, I too, like many others in this thread, recommend "The Mythical Man Month". However, I think, it's more important to figure out what one wants from life than to just better one's craft. Of course if that is what you want in life, then indeed pursuing that makes sense. I'm also tempted to recommend "Siddhartha" by Herman Hesse and "The Consolations of Philosophy" by Alain de Botton.

> recommend "The Mythical Man Month"

But be prepared to come away very depressed than a book that's now 44 years old contains wisdom that none of the people who ought to have absorbed it have.

If we're going meta, I'd add Ecclesiastes (yes, from the Bible). Doesn't matter whether you're religious--this was also a favorite of noted atheist Christopher Hitchens, for example.

The basic theme is that everything is ephemeral. This really starts to resonate when you see how fast the results of your labors over a period of decades fade, turn to dust, and are utterly forgotten. As you yourself will be in short order. Ultimate medicine for keeping things in perspective.

Here's a particularly readable paraphrase: https://www.amazon.com/Nothing-New-Under-Sun-Ecclesiastes/dp...

1. The Mythical Man Month. Quite old, but still a classic.

2. Dilbert. This comic series is an excellent introduction into how corporations actually work.

3. The Systems Bible. Somewhat obscure. Fascinating ideas on how large systems, both technical and organizational, actually work.

Dilbert! Amen. Humans are not Vulcans.

Peopleware: Productive Projects and Teams (https://www.amazon.com/Peopleware-Productive-Projects-Teams-...)

It talks about what makes a good team, how powerful they can be, and also gives some insight on team dysfunction.

You start career. If you already have been programming like in high school or something and finished CS university don't focus much on technical books. Learn how to work with people not computers and listen. I assume you know objects, ifs, for loops. It is enough for technical start. I don't expect more from junior. "The pragmatic programmer" is quite good start on it.

If you lack technical background and have problems with structure because you have not worked through university/school in structured way then "Code complete", "Clean code". Pick also something in depth for your specific area you are going to start, for C# good would be "C# in depth" to get really details of tech you will be working on. Get details in one programming language really well then you can pick up some other tech.

Le Petit Prince. Certainly you’ve read this but it should be reread on occasion and starting a career is a good time.

eg “You become responsible, forever, for what you have tamed”

How to Win Friends and Influence People

My favorites so far:

- Pragmatic Thinking & Learning by Andy Hunt

- The Design of Everyday Things by Don Norman

- The Effective Engineer* by Edmond Lau

- The Pragmatic Programmer by Andy Hunt and Dave Thomas

- SQL Performance Explained by Markus Winand

And http://teachyourselfcs.com curriculum is pretty good as well.

As a side note, there's so much we can learn from game programmers and OS/database/compiler engineers.

Anything about Complex Adaptive Systems. Holland’s short intro is nice.

Eventually you’re going to be faced with a very complex system to refactor. It helps to understand the principles of healthy complex systems in nature.


The Practice of Programming by K & R.

Or has this fallen out of favor?

It is by K&P, not K&R. Great book though.

This will be an unusual suggestion, but I highly recommend Geographic Databases and Information Systems by Stefanakis. The book is split into three parts, and the last one is specific to GIS. The other two, however, are a great introduction for databases, indices, data structures and even some graph theory. I've revisited several concepts I'd seen before at college.

Back when I was just graduating from college I picked up a book called "Coder to Developer: Tools and Strategies for Delivering Your Software". It was basically written for people who knew how to write code, but not all the stuff that comes with it in a professional setting - unit testing, build systems, version control, deployment and monitoring, etc.

I'd have a hard time recommending it per se, because it relies heavily for examples on specific technologies that were dated even when I was reading it half a decade ago, but I think something like it with that kind of content is a must-read for anyone starting in SE, especially coming from a CS degree that might have been light on those areas. If I could find a version that's more recent I'd recommend it to every new grad I know.

Find a book that introduces to programming patterns. I read 'Design Patterns' from the Head-First series when starting out http://www.headfirstlabs.com/books/hfdp/

The Art of Unix Programming. It teaches the "why"s, not just the "how"s.

This might sound either obvious or naive but if you are not a CS major, I would really emphasize reading CS textbooks such as CLRS, Programming Pearls, Dragonbook and most importantly textbooks on statistics m probability, linear algebra. There is a pyramid of skills for developers. People who haven’t studied throughly above stuff usually ends up at bottom (aka “enterprise” or “web” or “database” developers). They spend their lives in trivialities of patterns and plumbing and only very rarely doing something interesting such as distributed computing, game programming, systems programming, ML/AI.

For the technical side of things, http://teachyourselfcs.com gives a nice curriculum.

"Introduction to Algorithms" by Cormen et al. is one of my favorite books. It gave me a deep understanding of algorithms and data structures.

How do you compare it with Steven Skiena book?

Never had the chance of studying it. I'll give it a closer look. Thanks.

Most books recommended here are for programming of some or administration, anyone have notable books for cyber security, pen testing, red teaming, etc?

+1, that would be awesome! :)

1. Mythical Man Month - obvious reasons

2. Deathmarch: The Complete Software Developer's Guide to Surviving 'Mission Impossible' Projects- also a classic.

3. Thinking fast and slow - a great introduction to biases that will affect your thinking in any project.

4. The Art of War - strategy and tactics for attacking any problem. My personal favorite translation is by Thomas Cleary.

Don't read "The Design of Design" though. Brooks had some amazing insights in The Mythical Man Month and managed to be clear an concise in the points he wanted to make.

The Design of Design is just Brooks building a house. How the building project relates to software and what his point is, is never really made clear.

It's next to impossible to find a soft copy of "Deathmarch", except for in Russian. I tried but I haven't found yet.

Do you mean this one? https://www.amazon.com/dp/013143635X ? Or do you mean a different book?

Yes, the link is about a hard copy, not a soft one.

Is hard copy vs soft copy different than hard cover vs soft cover?

Yes. Hard copy is printed (hard or soft cover); soft copy is electronic.

Huh, I’ve never heard “soft copy” used for that before—I’ve always used electronic copy or digital copy. TIL.

Something for any industry that interests you. Computers are a tool. We use them not because a bubble sort is n-squared and quick sort is nlogn with a n-squared worst case. Step outside of your box and learn something about the real problems in the world and you will be the better as you realize how you can apply computers to solve the real problems.

I'm surprised to see basically only books on how to be better at making software. Nothing on negotiation or other soft skills like that.

Do people generally just not value things like negotiating? Or do you just not think a book would have helped younger you?

I can recommend two books, Crucial Conversations, and Getting to Yes.

Soft skills are irrelevant for programmers.

The Little Schemer

The Principles of Product Development Flow: Second Generation Lean Product Development

It changed my understanding and thinking about how best to work flow development. I highly highly recommend this book to coders to managers.

- The Pragmatic Programmer

- Refactoring

- Facts and Fallacies of Software Engineering

- Code Complete

Code Complete. I found Code Complete to be as eye opening for day to day work as SICP is while learning in University.

The title makes me wonder, is it too late to read these books for me, a decade long programmer?


Linux kernel coding standard, for attitude rather than specific guidance?

Waltzing With Bears. Oh my god how I wish I had read that one book.

(1) Object Thinking by David West. (2) Thinking Forth by Leo Brodie

Code Complete 2. I bought it... never read it.

Effective Engineer by Edmond Lau

Game Programming Patterns

Web Scalability for Startup Engineers

The Glass Bead Game by Herman Hesse.

1) The Art of Computer Programming

2) The Art of Computer Programming

3) The Art of Computer Programming

4A) The Art of Computer Programming

Has anyone really read these books from page to page?

Also, are they really relevant when you are starting your career? These seem more into the category of making you competent in the field of algorithms research, rather than something a person starting a career should read.

> Has anyone really read these books from page to page?

I recently finished reading the first three, and have started on the fourth (with a detour to read "fascicle 1.1 on MMIX"). I started the series planning to work every single problem. That turned out to be just a little unrealistically ambitious - I settled for _attempting_ every single problem except for the "unsolved research problems" (in some cases, just figuring out what the problem was asking for left me with a sense of accomplishment). Even so, I ended up spending a year on each volume, reading/working problems for a half hour a day or so. On the one hand, I loved these books, and really enjoyed reading/working through them, but - recommended before beginning one's software development career? You'll need to know a fair amount of calculus to make sense of a lot of it, and it'll take quite a while to get through. It's questionable how relevant it really is, fascinating as the subject matter is.

I've read 3 pages and I started on one exercise: I'm the better for it. Someday I hope to find the energy to read 3 more pages. If I'm really energetic I'll attempt another exercise as well.

The book everyone recommends but few have actually read.

Hackers and Painters

The Phoenix project. Everyone should read, IMO. I actually started with The Goal first. Then Phoenix project and then the Google SRE book.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact