Other less technical but relevant skills:
Linux (which evolved from UNIX) is the tool for the most important jobs of 2022. Even though UNIX dates back to the 70s, learning to break even common problems and utilities into composable bash scripts like a UNIX boomer could be the best change in mindset you’ll achieve in the new year. Your productivity will increase in ways you cannot foresee and you will appear to be using magic to those stuck using graphics user interfaces.
You can’t see it, but UNIX and Linux are the underlying operating systems on almost every device in the world including your phone and pretty much every web site you have ever visited. This is why the fastest growing and highest paying jobs currently require at least fundamental knowledge of the Linux bash command line. The better you get, the more you will outshine every other candidate for any tech job.
Why bash shell specifically?
Bash is the most important programming language you will ever learn. Every line you type on the command line is programming, in bash. Combine several of those lines together into a file and you have a script. Unlike other programming languages bash empowers everything you do every time you use the command line, including coding in other languages. It is not enough to know how to code, bash command line skills allow you to test, deliver, and share your code by connecting and automating with other tools.
Learning the bash command line also specifically includes a strong understanding of the UNIX philosophy and use of filters, programs that accept input, transform or operate on it, and produce output. Mastering this approach allows reuse of filter utility scripts by chaining them together into pipelines and even using pipelines within your command-line text editor to write code that writes for you.
Bash is the default shell on 90% of Linux systems. It is also the safest
shell in existence. Bash also happens to be the most powerful shell you
can learn doing things that no other shell can (including completion and
pretty prompts that are popular with users of other shells). Learning
shellcheck with Bash ensures you can quickly make POSIX compatible
shell scripts as well when required.
How do you learn bash?
Read the manual. Seriously, man bash is by far the most comprehensive
way to learn it. There is not a book on the planet that covers the
material better. Perhaps start out more simply by researching how to
complete challenges first. But, for this language you should always go
to the manual page because there is just too much shit on the Internet
that is just dead wrong, inefficient, and dangerous at worst.
For example, StackExchange is sites like it are full of unnecessary and
stupid use of awk, sed, cut, tr and other subshells. These
“solutions” also frequently use the ancient single bracket conditions
([]) instead of the safe double-bracket alternatives. If you see
single brackets in bash code you know for sure the author had not idea
how to code bash. What’s worse, is that often the people posting this
shit are as pretentious as they are clueless. You don’t have fight. Just
read the manual page be better and safer and ‘always’ run any code
through shellcheck.
⚠️ POSIX shell scripts are inherently more dangerous and easy to fuck up than good bash. The idea that POSIX shell is somehow safer and more compatible is simply a myth in 2022 and can easily be objectively proven false.
“The Network is the computer.” (Sun, 1990s, now bankrupt)
Okay, it really isn’t (despite the insanity of the latest 90s) but the network is definitely so important and ubiquitous that understanding how the Internet actually works is not only mandatory but the more you learn the more it sets you apart from everyone else in very significant ways.
Kubernetes and containers have upped the complexity and scope of networking. Once upon a time as a developer you could be good just understanding the absolute basics. These days you have to be able to “bridge” container networks just to access stuff on your own local system. Kubernetes has even more complexity for all the right reasons that you need to learn if you ever want to deploy your application cleanly and securely into the cloud, which is everything in 2022.
One of the best ways to learn networking is to setup your own simple physical home network using used computers or Raspberry Pis and then put a sniffer on one to inspect all the traffic as it is happening. This will give you greater than book knowledge about how networks work. Who knows, you might even catch unauthorized traffic and catch a hacker on your network (like I did when I first did this).
Want the highest paying, most challenging tech job of 2022? Become a Cloud Native Infrastructure Engineer. The industry has the “largest tech skills drought in history” because it cannot fill the roles of cloud native infrastructure engineers right now where the median salary is 146,000 dollars. My friend, a veteran, just landed a job over 300,000 to help companies setup their on-prem cloud native infrastructure (and I’ve been offered over 250,000 as well, even though I took another job for less).
Docker might have started it, but containers have gone far beyond Docker
at this point. Sure we still use the docker command for everything,
but containers are now the fundamental building block of most enterprise
IT infrastructure. Not knowing how to use and create containers is a
serious red flag to most IT employers today (although apparently not
enough to stop me from getting a job as an Infrastructure Engineer where
I then learned it).
Like former me, you may be thinking, “I don’t need to learn containers, I just need to learn how to create applications and infrastructure.” Like me, you’d be dead wrong. Modern applications are containerized from the very beginning. If you get a green-field project and don’t think immediately about how that application is going to be deployed as containers — most likely into Kubernetes — then you are just doing it wrong in 2022. If your current team does not understand the significance of that or the reason why start looking for another job today.
Containers and Kubernetes are not just for shitty, insecure NodeJS apps and “DevOps.” (Node is being replaced by Go REST APIs bottled in much safer, lighter, and faster “FROM SCRATCH” containers, btw.) Containers and Kubernetes are being used at large scale to deploy high-performance machine learning model creation, services, and testing. This stuff is already everywhere and is going to continue to explode in scope more than it already has.
Okay, I’m not including bash (the most important programming language you will ever learn) because I already wrote about it. But the next most important languages to learn in 2022 are Go, Python, and (of course) JavaScript (which kinda includes HTML and CSS that most would argue are not true programming languages).
Why these three? Mostly, because they are not only in strong demand but have a very bright future ahead. Let me explain that a bit.
Go was specifically written to address the highly-concurrent, frequently compiled, performance sensitive requirements of the world’s largest Silicon Valley company, Google. Rob Pike and Ken Thompson brought the world C, UNIX, Unicode, Plan 9 and several other things as well as the Go programming language, which they call “a modern C”. Every major application and tool in the cloud native landscape is written in Go, most importantly Docker and Kubernetes. Most companies are actively replacing legacy code in NodeJS and Java with Go. Here’s one statement from a relatively small but important modern company like Uber:
“While historically Uber has been mostly a Node.js and Python shop, the Go language is becoming the language of choice for building many of Uber Engineering’s new services. There is a lot of momentum behind Go at Uber, so if you’re passionate about Go as an expert or a beginner, we are hiring Go developers. Oh, the places you’ll Go!” (2016 )
Go static typing fulfills the needs of most large projects better than
any interpreted language (yes even TypeScript) and cross-compilation
makes is flexible for everything including ultra-lightweight “FROM
SCRATCH” containers that require a compiled language. Go’s batteries
included selection of network and encryption modules — not to mention
gRPC and ProtoBuf — make it the dominant pick for cloud native
applications development as well. Even if you just want to work in cloud
native operations learning to read Go will help you understand what the
applications are doing when the documentation fails you. For example, by
reading the source code of kubeadm you can get a perfect understanding
of every step necessary to create a Kubernetes cluster.
Python’s greatest contribution to the world is a really kick-ass C stub and encapsulation framework. People write simple Python code, then when they come to stuff that needs improvement, they write C code and create a stubbed library module to import. This is why Python gets all the credit for the innovations of PyTorch, Tensorflow, NumPy, SciPy, Pandas, and other C code masquerading as Python. Python also includes a full version of SQLite (simple database) and TK (for graphics). Even though Python suffers from serious distribution issues (and always has) it is still mandatory learning for anything involving numbers, math, machine learning, systems automation, and cybersecurity.
Related: