At the most basic level, a computer works only with zeros and ones, which make up what's called binary code. Each one or zero is called a bit (from binary digit).

Any data in a computer is represented simply as a sequence of bits, for example images, music, and text. The familiar numbers from the decimal system can also be represented in binary form.

• 0 → 0
• 1 → 1
• 2 → 10

How do you encode text?

A computer doesn't "understand" text. To work with letters and other characters, they also need to be turned into numbers. This is done using encodings, that is, tables in which each character corresponds to a specific number.

The simplest way is to number the letters, starting with 1.

• a → 1
• b → 2
• ...and so on up to z → 26

Now we can represent the word hello as a set of numbers.

h e l l o
↓ ↓ ↓ ↓ ↓
8 5 12 12 15

And good turns into this sequence.

g o o d
↓ ↓ ↓ ↓
7 15 15 4

The program doesn't know that this is a word. It simply sees the instruction "display the character with code 8, then the one with code 5, and so on".

ASCII. The first mass encoding

The first computers worked mainly with English. For it, the ASCII table was invented in the 1960s, which included 128 characters, among them the Latin alphabet, digits, punctuation marks, special characters (@, #, !, \n), and control codes.

This was enough for the first programs, but not for the whole world.

When computers began to be used in other countries, a problem arose. ASCII has no Cyrillic, hieroglyphs, Arabic script, accents, currency symbols, and so on.

Each country or company began to make its own encoding based on ASCII.

• Windows came up with Windows-1251 for Russian
• Apple created Mac Roman
• The countries of Eastern Europe, Asia, and the Middle East developed their own variants

All these encodings were incompatible with each other. Code 226 in one encoding could be the letter é, in another the letter и, and in a third a technical character altogether. This led to real chaos.

What encoding problems looked like

If you see this in text.

ÐÑивеÑ!

It means that the program determined the text's encoding incorrectly. It received a sequence of bytes but read them with the wrong table.

This was the norm in the 1990s and 2000s. One program wrote text in Windows-1251, another read it as ISO-8859-1, and the result was garbage.

Unicode and UTF-8. The end of the mess

To fix everything, in the 1990s people began to create the universal Unicode table, which contains characters of all the world's writing systems, among them the Latin alphabet and Cyrillic, Chinese and Arabic script, mathematical signs, ancient Egyptian writing, and even emoji.

Within Unicode there are several storage formats. The most widespread of them is UTF-8. It encodes English characters compactly but can expand to fit any others.

Today UTF-8 is the default standard on the internet, in Linux, in databases, and in code editors. PHP source code and strings in modern projects are also customarily stored in UTF-8.

Why does a programmer need to know this?

• You will work with text, and encoding errors still happen, especially when reading files, processing data, and interacting with APIs and databases.
• PHP stores a string as a sequence of bytes, so when working with files and databases the encoding sometimes has to be specified explicitly.
• You need to be able to diagnose problems. For example, if you see "garbled characters", it's almost certainly an encoding error.

The program receives the numeric codes of characters and displays them on the screen — this is handy when a character is hard to type on the keyboard. Find the characters with codes 126, 94, and 37 in the ASCII table (we're interested in the decimal code — dec or decimal) and print each one on a separate line using the chr() function and a line break:

print_r(chr(...) . "\n");
print_r(chr(...) . "\n");
print_r(chr(...) . "\n");

For example, the character ? has code 63:

print_r(chr(63)); // output: ?