Git turns a plain folder into the repository from yesterday. You run git init in a folder, and from then on Git can track it. You make changes, tell Git which ones to record, and it saves a snapshot with your name, the time, and a message. That snapshot is a commit, tomorrow's word. Git's history is a chain of those snapshots. You can compare any two, roll back to any one, or start a separate line of changes and merge it in later. Days 21 to 24 cover those pieces one at a time. Git is local by default. Nothing leaves your machine until you push to a remote like GitHub. That's why you can work on a plane and sync later. Quick check: Git and GitHub. Which one is the tool on your computer, and which one is the website? ## SPOKEN LINES - Two of you edit the same file on the same afternoon. Whoever saves last wins, and the other person's hour of work is gone. - - GIT - - Git is a version control system. It watches your project folder, and when you tell it to, it records a snapshot with a note about what changed. - - You can undo a change from six months ago, two people can edit the same project at once, and a risky idea can live in its own line and get thrown away. - - So Git is the system that tracks every change to your code. Tomorrow: Commit, the snapshot Git takes when you tell it to. ## NEXT Day 21: Commit

Yesterday's package needs somewhere to be developed. That's the repository. Think of it as a project folder with a memory. Every change is recorded. Nothing is lost when you overwrite a file, because the earlier version is still in the history. A repo is also the unit of sharing. Open source packages on npm and PyPI almost always point back to a public repo, usually on GitHub, where you can read the source and report bugs. Two copies is normal. The local repo on your machine, the remote repo on GitHub or GitLab. You work locally and push changes to the remote when they're ready. Tomorrow covers the tool that makes that work. Quick check: You delete a file from a repo by accident. Is it gone for good? ## SPOKEN LINES - Your first Monday on a new team, and nobody sends you the code. The only instruction is clone the repo. One folder shows up on your laptop, holding the whole app plus three years of changes. - - REPOSITORY - - A repository is the home folder of a project. It keeps the full history: every version, who changed what, and when. - - You can open a file the way it looked six months ago, see who wrote line 40 and why, and get everything back if your laptop dies. - - So a repository is where a project's code and its history live. Tomorrow: Git, the tool that writes that history. ## NEXT Day 20: Git

Three words from the last three days, and they overlap on purpose: A library is the code. A package is the code wrapped for distribution, with a name and a version. A dependency is a package your project has installed and relies on. A package has a small file that says what it's called, what version it is, and which other packages it needs. That's how the dependency tree from yesterday gets built: the package manager reads that file, fetches what's listed, then reads those files, and keeps going. Your own project is a package too. package.json at the top of a JavaScript project describes it the same way. Quick check: You run pip install requests. Which of the three words above describes what pip just downloaded? ## SPOKEN LINES - A client wants card payments by Friday and you have never touched payment code. You type npm install stripe, and seconds later your project can take a card. - - PACKAGE - - A package is code bundled up with a name, a version number and a list of what it needs, so a tool can download and install it. npm does that for JavaScript, pip does it for Python. - - Stripe handles the card payment, express handles the routes, and dayjs turns a timestamp into 3 minutes ago. - - So a package is code boxed up so it can be installed and reused. Tomorrow: Repository, the folder where your whole project and its history live. ## NEXT Day 19: Repository