This article is geared toward 3rd-year computer science students or people that have some basic programming experience. You should know about loops, conditionals, functions, exceptions, arrays, maps, and objects at a minimum. If you are a beginner, complete this course first before tackling the following material.
If you aren't in a computer science program, that's OK! There are plenty of free resources to learn computer science on your own. I will link to online courses and projects that you can complete to learn computer science on your own. The resources and projects resemble what you would find in a university computer science program.
It doesn't matter what part of the stack you are working on, software engineers need to develop data structures to represent models of the world. In this course, you will learn what it means to create abstractions, and you will learn about the abstractions that are commonly used to solve problems. A course on data structures and algorithms will teach you
- basic structures like lists, stacks, and queues
- more complex structures like trees and graphs
- basic sorting and searching algorithms
This course directly teaches skills that are often tested during software engineer interviews, so strictly from a getting-a-job perspective, this course is extremely valuable.
For learning about data structures and algorithms, check out Princeton's course. This course will give you experience applying algorithms to real-world problems. Once you're comfortable with the basics, Leetcode has an endless supply of programming problems you can practice to your heart's content.
Not sure what a software engineering interview is like? Here is an example provided by Google.
You've written the perfect solution to your program. How do you know your algorithm is correct? You can write unit tests, but how do you know the algorithm is correct for all possible inputs? For this, you would need to prove your algorithm is correct. A course on analysis of algorithms will teach you how to design algorithms and rigorously prove their correctness. This course will also take your ability to solve interview problems to the next level. Expect to learn
- graph theory
- greedy algorithms
- dynamic programming
- max flow, min cut
Check out Stanford's course on algorithms to get you started.
The goal of a course on programming languages is not to simply teach new languages. The goal is to teach about programming languages. Programming languages come and go. But knowing programming paradigms allows programmers to quickly pick up new languages because they are familiar with the underlying concepts. A course on programming languages will teach you about
- paradigms such as functional and object-oriented
- dynamically-typed and statically-typed languages
- idioms like closures, lexical scope, and mutation.
To learn programming languages, check out the University of Washington's course.
Without hardware, there is no software. A course in computer architecture will give you a valuable perspective when working on software. For instance, you'll learn that disk is slower than main memory, and main memory is slower than an L1 cache. You'll learn that performance optimizations in hardware have even led to some critical security vulnerabilities, like Meldown and Spectre. You'll understand why we use GPUs to process matrices of data for machine learning. In a course on computer architecture, expect to learn about
- assembly and machine code
- memory hierarchy
- CPU design patterns such as pipelining, branch prediction, and hardware register renaming
An operating system is software that sits in the middle of your applications and your computer architecture. An OS is responsible for a lot of things, like concurrently running all your applications seamlessly, storing your files on disk, and handling I/O. An OS is naturally a complex piece of software in order to wear all of these hats. Expect a course on operating systems to teach you
- process management
- synchronization of threads
- traps, exceptions, and interrupts
- basics of file systems
Let's say you've written a program in Java. But your computer doesn't understand Java. How do you turn your code from the Java language to a language the computer understands? There is a piece of software called a compiler for that.
A course on compilers will give you a deeper understanding of programming
languages and how they are implemented. It will demystify what is happening when
you hit the big green button in your IDE or type
javac in your shell. A course
on compilers will teach you about
- regular expressions
- abstract syntax trees
- code semantics, optimization, and generation
The internet is kind of a big deal. I mean, without it, you probably wouldn't be reading this article. But how do computers talk to each other? In a course on networks, you will learn how computers communicate over the network. You will learn about the technologies that are prevalent in today's networks, such as IP, TCP, UDP, HTTP, and apparently, a lot of acronyms that end with P.
Most software services need to be able to communicate over the network in some way, so having networking knowledge can come in handy.
Check out MIT's course to learn about computer networks.
Google allegedly processes 5.8 billion search queries per day or 70,000 per second. In order to serve billions of users at any time of the day, we need to do what is called "scale horizontally." A single computer can only be so fast and powerful. To address this, engineers coordinate many computers to talk over the network in order to reliability serve many customers simultaneously.
For learning distributed systems, do the dslabs projects. You will get hands-on experience developing a fault-tolerant, highly available, sharded key-value store. Supplement the labs with these lectures from MIT.
Big data is a major part of society now. We need systems to store and query larges amounts of data. Many software systems utilize a database in some form or another. Knowing about how these systems work can help you when designing database schemas for your own applications. Expect a course in database systems to teach you
- the SQL query language
- hash and B-tree indexes
- ACID transactions
- write-ahead logging
To learn databases, complete the SimpleDB project, where you will implement a SQL database that supports B-tree indexes, transactions, write-ahead logging, and query optimization. Use these lab specifications to guide you.
Name a piece of software that is not at threat by attackers. Most (all?) applications or software services are at risk of being attacked by bad actors. In a course on security, you would learn about common threat models and what you can do as a programmer to avoid them.
Check out the University of Helsinki's course to get you started.
Technology is always evolving, so it's hard to make a list of only 10 courses. So, here are 2 more bonus courses for you to consider.
Making your app accessible and easy to use is perhaps just as important as the underlying software that drives it. In a course on HCI, you learn how to design novel interfaces that can be applied to everything from web and mobile apps, to point-of-sale systems and ATM machines. A course on HCI will teach you principles of designing usable and accessible interfaces.
Machine learning is the technology that recommends videos on YouTube, serves personalized ads, self-drives cars, and guesses "hot dog" or "not hot dog". Spoiler alert: It's math. In particular,
- linear algebra
If you want to go deep designing novel machine learning algorithms, you should be well equipped mathematically. Professor Leonard is a great resource to get you started learning math.
For additional information and guidance on your path to learning computer science, check out Teach Yourself Computer Science.