🐍 Start Python on a MacWindows · เริ่มเขียน Python บน MacWindows · 在 MacWindows 上开始学 Python

🇬🇧 English

By the end of this page you will be able to sit down at a Maca Windows PC, open the Terminal, create a Python file with your own hands, and run it. Nothing is assumed. We start with the computer itself.

You need a Maca Windows PC, about two hours, and patience. Every step has a command you can copy.

The path: your MacPC → files and folders → the Terminal → install Python → write a file with vim → run it → IDLE → Jupyter → two full lessons in a notebook.

🇬🇧 English

Python was written by one man. Guido van Rossum, a Dutch programmer, started it over the Christmas holiday of 1989. His office was shut and he was bored.

He led the project for nearly thirty years. Python users called him the Benevolent Dictator For Life — a joke title, but he really did settle every argument. He stepped down in 2018.

And the name is not the snake. He was reading the scripts of Monty Python's Flying Circus, a British comedy show, and wanted a name that was short and a little silly. The snakes came later, on the logo.

This matters more than it sounds. It is why Python's own documentation is funny, and why the language values code you can read over code that looks clever.

🇬🇧 English

Programming is taught in English, so the words come first. Learn these twelve and most instructions stop being mysterious.

Notice that many of them are ordinary English words doing a special job. A file is a file. To run something is to make it go. The words are not hard — they are just precise.

Click any card to hear it said aloud.

🇬🇧 English

Below is a real working terminal, inside this page. Nothing you type can break anything. Type help to see every command it knows.

Almost everything here works on both systems. Windows PowerShell understands pwd, ls, cd, mkdir, cat and rm — they are nicknames for its own longer commands.

The practice terminal below draws a Mac-style prompt (student@mac:~$). On Windows yours will look like PS C:\Users\You> instead — the commands you type are the same.

Two do not: touch and grep. On Windows use New-Item notes.txt and Select-String instead. The simulator accepts the Mac names so you can practise the idea either way.

🇬🇧 English

grep Chiang notes.txt looks for the exact word. But the text you give grep is not a plain word — it is a pattern, and patterns can describe a whole family of words at once.

These patterns are called regular expressions, or regex. A few characters have special meanings. Learn these eight and you can read most of the regexes you will meet.

The same patterns work in Python, in your editor's Find box, and in almost every programming language. Learning them once pays you back everywhere.

🇬🇧 English

Type python3python on its own and press Return. The prompt changes to >>>. You are now inside Python.

Everything you type is run immediately and the answer printed. This is called the REPL — Read, Evaluate, Print, Loop. It is the best place to test one small idea.

To leave, type exit() and press Return, or press CtrlD. The prompt goes back to $.

Know which prompt you are looking at. $ means the Terminal is waiting for a Terminal command. >>> means Python is waiting for Python. Typing ls at >>> gives an error, and that error confuses more beginners than anything else on this page.

🇬🇧 English

The REPL forgets everything when you leave it. A script is a file, so it lasts. This is how real programs are written.

Three steps, every time:

1. cd into the folder holding the file.
2. python3python hello.py
3. Read whatever comes back.

The most common error by far:

python3: can't open file 'hello.py': [Errno 2] No such file or directory

This almost never means the file is gone. It means you are in the wrong folder. Type ls. If you cannot see hello.py in the list, Python cannot either. cd to where it actually is.

🇬🇧 English

IDLE comes free with Python. It is a small editor plus a REPL in one window, with no setup at all. Start it from the Terminal:

Two windows matter. The Shell window is the same >>> prompt you already met. File → New File gives you an editor; write code there, press F5 to save and run it, and the output appears in the Shell.

IDLE is worth thirty minutes because it removes the Terminal from the loop while you are still learning the language itself. It colours your code and points at syntax errors. Move on when it starts to feel small.

🇬🇧 English

A notebook is a document that mixes writing and code. The code sits in cells you can run one at a time, and the answer appears directly under each cell. It is the best way to learn a language, because you can change one line and see what breaks.

Try it before you install anything. Go to jupyter.org/try and pick JupyterLab. It runs in your browser, on their computer, with nothing to set up. Press ShiftReturn on a cell and watch it run. Your work is not saved there, so it is for playing, not for the lessons.

When you are ready to keep your work, install it once, then start it from the folder holding your notebooks. Your browser opens by itself.

The three things to know: ShiftReturn runs a cell and moves on. Cells share memory, so a variable made in cell 3 still exists in cell 9. Run cells in order — running them out of order is the cause of almost every strange result.

Stop the server with CtrlC twice in the Terminal window that started it.

🇬🇧 English

A notebook that only exists on your laptop is one spilled drink away from gone. GitHub keeps a copy online, remembers every version, and lets you show your work to a teacher or an employer.

You need a free GitHub account first. Go to github.com/join and sign up yourself — never let a program create an account for you.

Then install git and tell it who you are. You do this once, ever.

Clear your output before you commit. A notebook stores its results — every number, every picture — inside the file. That makes the file huge, makes every change look enormous, and can publish data you did not mean to share. In JupyterLab: Kernel → Restart Kernel and Clear Outputs of All Cells, then save, then commit.

🇬🇧 English

GitHub stopped accepting account passwords over HTTPS in 2021. Instead you make a personal access token — a long random password, made for one computer and one job, that you can cancel at any time without changing your real password.

Make one like this:

1. On github.com, click your profile photo, top right → Settings.
2. At the bottom of the left sidebar → Developer settings.
3. Under Personal access tokens → Fine-grained tokens → Generate new token.
4. Token name: something you will recognise later, like MacBook — python lessons.
5. Expiration: 90 days. Do not choose "no expiration".
6. Repository access: Only select repositories, then choose your python-lessons repo.
7. Permissions → Repository permissions → Contents → Read and write. That single permission is what git push needs.
8. Click Generate token, then copy it immediately. GitHub shows it once and never again.

Now run git push. When it asks for your username, type your GitHub username. When it asks for your password, paste the token. Nothing appears on screen while you paste — that is normal, not a broken keyboard.

🇬🇧 English

A token is a password. Treat it like one:

• Never type it into a file, a notebook cell, or a screenshot.
• Never email it or paste it into a chat, not even to a teacher.
• If it ever leaks — or you commit it by accident — go back to the same page and click Revoke, then make a new one. Revoking is instant and costs you nothing.
• This is the other reason to clear notebook outputs before committing: an output cell can quietly contain a token you printed while testing.

🇬🇧 English

Read the three commands as a sentence: add chooses what to save, commit saves it with a note about why, and push sends it to GitHub. Nothing leaves your computer until you push.

Write commit messages a human can use. "Finish section 5 exercises" tells you something in a year. "update" does not.

Git is a big subject and this is the smallest useful piece of it. When you want the rest — branches, pull requests, undoing things — work through Introduction to Git, then the GitHub Pages lab.

🇬🇧 English

This is the language itself, in ten short sections. Download it, put it in ~/Documents/Code, and open it with jupyter lab.

Every section now ends with an exercise and a worked solution. Do the exercise before you open the solution. Reading a solution feels like learning and is not.

Expect it to take two or three sittings. If a cell errors, read the last line of the error first — it names the problem.

🇬🇧 English

This is what turns "I can read Python" into "I can write a small useful script". Four topics, each of which you will use in every real program you ever write.

Section 2 writes real files into whatever folder Jupyter started in — so start it from ~/Documents/Code and your files land somewhere you can find them.

Section 3 has one cell that fails on purpose. That is not a mistake in the notebook. Read the red traceback carefully; learning to read errors is the single most useful skill on this page.

🇬🇧 English

Say you are writing the space game in the next lesson. Each enemy — each creep — needs a speed, some hit points, and a path. Without classes you get this:

Two creeps, six variables. Twenty creeps, sixty variables. And nothing joins creep1_speed to creep1_hit_points except the name you typed.

A class makes one creep into one thing. Then the dot reaches inside it.

Read the dot as the word of, backwards. creep.speed is the speed of this creep. On the left, which thing. On the right, which fact about it.

🇬🇧 English

Put all the creeps in one list. Now one loop moves every one of them — and each uses its own speed, because creep.speed always means this creep's speed.

Three lines handle three creeps or three hundred. You never match up index numbers, and you can never accidentally give creep 7 the speed of creep 3.

The dot reaches actions as well as facts. creep.update() is the update of this creep. Facts are called attributes; actions are called methods. Same dot for both.

🇬🇧 English

You have been using the dot all along. In Python almost everything is an object, and the dot is how you reach inside it. You have already written all of these:

🇬🇧 English

When both notebooks feel comfortable, the next lesson is Python Toolkit. It covers the seven tools you reach for after the basics: classes, generators, decorators, type hints, pathlib, json and requests. It has its own notebook with eight more exercises.

Then build something. Make a Space Game and Map Chiang Mai are both step-by-step projects that use the classes you just met. Grokking Algorithms takes you further, through sixteen classic algorithms with their own notebooks.

When you want to check what stuck, take the Python Exam. It only asks about things taught on this page and in the two notebooks.

After that: install VS Code (brew install --cask visual-studio-codewinget install Microsoft.VisualStudioCode) and write something you actually want. A program you need beats a tutorial you finish.