KB: Git HEAD explained using the "Tape Head" Analogy

 

The “Tape Head” Analogy for Git HEAD version control

 

HEAD = The read/write head on a cassette tape

Think of your entire Git repository like an old cassette tape with multiple tracks (branches).

✔️ HEAD is the physical read/write head.

It doesn’t store music itself — it just points to where on the tape you want to read or write.

✔️ Switching branches = moving the tape head

When you run:

git switch feature/login

You’re not changing the branch itself —
you’re just moving the tape head to read/write on that track.

HEAD only exists once

There is one tape head, but many tracks.
The head points to whichever track you choose.

---

Branches = multiple tracks on the same cassette

Each branch is like its own “audio track” that ends at a different timestamp (commit):

main -> Commit A
feature/login -> Commit C
bugfix -> Commit F

All those tracks exist, but the HEAD can point only to ONE at a time.

---

Commits = positions on the track

When you rewind to an old commit:

git checkout 2a9e4d

HEAD now points directly to a track position, not a branch name.

This is called a detached HEAD,
like moving the tape head to the middle of the tape, not to the end of a track.

---

Commit = writing new audio where the head is pointing

When you commit:

• You’re recording new audio at the location HEAD is pointing to.

• If HEAD is pointing to a branch → that branch advances.

• If HEAD is pointing to a commit (detached) → Git creates a temporary new history.

---

Summary:

HEAD is not part of the branches themselves.
It’s a mechanical tool inside my Git working directory.
Whenever I switch or checkout, I reposition the HEAD to point to a branch or commit.
All Git actions (read, write, commit) happen through that HEAD pointer.
HEAD is just the “tape head,” and the branches/commits are the tape itself.