Programming fundamentals

Every language dresses the same few ideas in different syntax. A value is stored under a name; a condition chooses between paths; a loop repeats until something changes; a function packages steps so they can be reused and checked; a collection holds many values in a shape suited to the questions you will ask of it. SFIA calls the resulting skill programming and software development, the building of software components, and assumes these ideas are settled before any of its higher levels begin. They are the grammar. Every framework is vocabulary built on top.

The machine under the language

Fundamentals also include enough of the surrounding computer to stop it feeling like magic. MDN's getting-started material deals with environment setup, the file system and the command line before any page is written, and that order suits every kind of programming. You need to know where a file is saved, which program runs it, how the code you wrote differs from the output it produced, and why one script behaves differently in two folders. Whether a language is compiled ahead of time or interpreted as it runs, whether a variable holds a value or a reference to one, and what an exception does as it climbs out of a function: these questions arrive in the first month and never stop arriving.

• Data structures chosen on purpose. A list when order matters, a dictionary or map when you look things up by key, a set when all you ask is whether something is present. The wrong choice is the commonest cause of slow, tangled beginner code.
• Debugging as a method. Read the whole error message, go to the line it names, form a guess, change one thing at a time. Printing values is a legitimate tool; a debugger that pauses the program mid-run is a better one.
• Reading other people's code. Most working hours go on existing code, and following an unfamiliar function through three files is a skill that needs practice of its own.
• Decomposition. Breaking a vague task into steps small enough to write and check one at a time, which is the part no syntax reference can teach.

Fundamentals do not need a particular language, but they need one held long enough to stop fighting its syntax. Python and JavaScript are the usual picks because both run with little setup. C, which Harvard's CS50x opens with, exposes more of the machine at the price of more friction. Cisco's free Python Essentials 1 is graded on code that actually runs, which makes it an honest self-check. From here the field opens up: programming languages is the question of which one to be hired in, and software engineering practice is what turns working code into code someone else can maintain.