Knowledge · IT · Data · DX

Git · GitHub

Git / GitHub

📂IT · Data · DX⏱2-min read📊Practitioner🔗4 related

What is Git?

In one line
Git is a version control tool that stores and manages the change history of code; GitHub is the leading service for hosting that code online and collaborating on it.
Bring this topic to your team📚 Related Courses💬 Ask About Training
⚡ Summary in 3 lines
  • Git = tracks code changes
  • GitHub = online storage and collaboration
  • Foundation for rollback and collaboration

Adopting Git changes how a team manages files. Instead of piling up copies like 'app_final.py' and 'app_final_really_final.py', you keep a single file and the change history stacks up beneath it. Who changed which line, when, and why is recorded alongside the commit message, so when something goes wrong, it becomes a matter of looking up the record, not assigning blame. When you hand code you wrote alone over to someone else, the history itself serves as the handover document.

Git and GitHub are often mentioned as if they were one word, but they sit at different layers. Git is a program that runs on your laptop; GitHub is the server where you store its output. Many companies use GitLab or Bitbucket in the same role. In practice you need four concepts: commit (a save point), branch (a separate line of work that leaves the main line untouched), pull request (a request for review before merging), and merge (combining the work). You must choose between a public and a private repository, each file is limited to 100MB, and binaries such as Excel files and images can't be compared change by change, so it's better to keep mostly code and configuration files in it.

When you build code with AI, the commit is your unit of experiment. If you ask for five features at once and leave a single commit, you can't tell which request caused the problem, so it's safer to follow the rule of one request, one commit. A GitHub repository can also connect directly to a deployment service, so the live site updates the moment code is pushed. In that setup, pushing is deploying, so build the habit of never pushing without review.

The most common mistake in the field is committing code with an API key or database password written in it, then pushing it to a public repository. Even if you delete it from the file later, it stays in past commits, and a few lines of history search will bring it back. The second misconception is 'It's on GitHub, so the backup is done.' If you commit but never push, that record still lives only on your laptop and disappears if the device fails. The third is filling every commit message with 'fix': if you can't find the point to roll back to, the history you built up is as good as nonexistent.

Courses that cover this term · AI & DX
Recommended courses · Recommended instructors · Quick quote
🧮 Mini quote · Based on instructor fees
About KRW 1.2 million (≈ US$900)2-hour talk · up to 20 people · Korean
Get a quote for these requirements →
💡 In plain English
Git is an 'edit history with undo' for text, and GitHub is like sharing on Google Docs — the basics of collaboration and version control.
Why it matters in practice
You can experiment with AI-generated code freely only if you can undo a mistake. Git is that “undo” safety net.
Origin and sources

Git was created in 2005 by Linus Torvalds, the creator of Linux, for developing the Linux kernel. GitHub launched in 2008 and became the world's largest platform for code collaboration.

Source · Linus Torvalds, Git (2005) · GitHub (2008) · Read the original
An example

Consider a hypothetical case: Company A, an industrial parts distributor with 38 employees and annual revenue of about KRW 12 billion (≈ US$8,978,000). Kim, the inventory manager, had been building a 'warehouse inventory lookup web app' alone for three weeks through vibe coding, and the copies on their desktop had grown to 27. One day they accidentally overwrote a working version and lost two days of work.

  1. The day after the incident, Mr. Kim told Mr. Park, the IT manager, 'If the AI rewrites the whole codebase, there's no way to undo it.' Mr. Park installed Git on his laptop and spent an hour teaching him nothing but how to make commits.
  2. Every time the screen worked properly, Kim left a commit with a message like 'Warehouse inventory list lookup works', and in the first week alone there were 41 save points.
  3. When asking AI to 'add an Excel export feature' also broke the lookup screen, two commands rolled the project back to a commit from 20 minutes earlier, so no work was lost.
  4. They agreed to create a private repository on GitHub and push before leaving each day. But Mr. Park first spotted the warehouse database password hard-coded in the source, and only made the first upload after moving it into an environment variable.
  5. Three months later, Ms. Lee from Sales joined and worked on separate branches. Time spent on recovery fell from 4 hours a week to 20 minutes, and turnaround on change requests dropped from 5 days to 1 day on average.
Related terms
Training linked to this term
📚Building a Real-World Website with Claude CodeNanum Consulting corporate training · On-site/OnlineView Course →
Need in-house training on this topic?
Get a curriculum and quote tailored to your needs right away.
Ask About Training
Last updated 13 Sep 2026 · Reviewed by John Kim

FAQ

No. Git is a 'tool'; GitHub is a 'service' for working with Git repositories online.
If you do vibe coding, just knowing how to roll back to a commit is a big safety net. Most AI tools handle it for you.
Git is a 'version control tool'; GitHub is the 'service' where that code is stored online and teams collaborate on it.
Useful for collaboration and history tracking, so more non-developers, such as document writers and planners, are using it.