Installing Git & first commit

You know why version control exists now. Here's where it gets a name, a face, and your first real fight with a teammate.

You open a terminal on Rachit's laptop. First problem, before you've even made a single commit: Git doesn't come installed by default on most machines, and even once it is, it doesn't know who any of you are.

Getting Git onto the machine

Pick your platform:

Good. Git exists on the machine now. It still knows absolutely nothing about who Rachit is.

Here's a scenario that happens in the design lab downstairs every semester. There's a sign-up sheet on the door for booking lab time. It has one column: a time slot. No name column. Two people book 4 PM by writing "4–5 PM" on two different lines, and when there's a clash at 4:03, nobody can tell who has the actual claim — the sheet recorded that a booking happened, but not whose.

A commit without an identity is that sheet. Git will happily record a snapshot with zero information about who made it, and six months from now, when someone asks "wait, why does the hero section have this weird padding — who added it and can I ask them?", the answer would be a shrug. This is exactly the "you don't know when or why" failure from Chapter 1, except now it's "you don't know who." So before the first commit, every one of you sets an identity — once, on your own machine, and Git remembers it for every project you ever touch there:

Turning Rachit's folder into a repository

Now the actual moment. You're standing behind Rachit; he's got the folder open in the terminal.

Nothing visibly changed. That's the point — git init just created one hidden folder, .git/, sitting quietly next to index.html. Every snapshot this project will ever have is going to live inside that folder. Delete it, and the project reverts to being just files with no memory. It's the difference between a house and a house with a foundation poured underneath it — from the street it looks the same, but only one of them can hold up a second floor.

git status is the command you will type more than any other for the rest of your life with Git. It costs nothing and it never lies. "Untracked" means: I see these files exist, but I've recorded nothing about them yet.

That's it. Forty minutes of Rachit's work is no longer just sitting on one laptop as loose files — it's a permanent, timestamped, attributed record. git add tells Git "these are the files I want in the next snapshot" (more on why this is a separate step from committing in Chapter 4). git commit takes the snapshot and signs it with the identity you configured a minute ago.

Keeping the noise out: .gitignore

Anvesha's about to start on content, which means she's about to run npm install for some Markdown-rendering package, which means a folder called node_modules is about to appear with roughly four thousand files in it that nobody wrote and nobody wants tracked. If she runs git add . blindly, all four thousand land in the repo, and every future git status becomes unreadable.

A file named .gitignore, sitting at the root of the project, tells Git which paths to never even mention:

# Compiled output / dependencies
node_modules/
dist/

# Editor cruft
.vscode/
.idea/

# Secrets — never commit these
.env
.env.local

# OS junk
.DS_Store
Thumbs.db

The .gitignore file itself gets committed — that's how the whole team ignores the same things without each of you configuring it separately.

fix: hero padding

adjust

fix(hero): remove double padding on mobile breakpoint

The hero container and its inner wrapper both had 16px
horizontal padding below 480px, stacking to 32px total and
pushing the CTA button off-screen on small phones.

Tested on iPhone SE simulator and a real Redmi.