Systems / Git
Build a real commit, byte by byte.
BY THE END, YOU CANExplain what Git stores when you commit, and why branching is almost free.
01 / Try the idea
Type anything into the file. These are the object IDs Git would store — run the command underneath to check them against Git on your own machine.
6 bytes → hashed as blob 6\0…
hashing…hashing…hashing……Verify the blob yourself: printf "hello\n" | git hash-object --stdin
02 / Check your understanding
Read the idea in plain text
The file containing `hello\n` is six bytes. Git hashes the string `blob 6\0hello\n` and gets ce013625030ba8dba906f756967f9e9ca394464a. A tree naming that blob as greeting.txt hashes to 57e9529754dc514a3ec10db2ff882018fbe1fcbf, and a commit pointing at that tree, authored by Ada at a fixed timestamp, hashes to c53d2a0d16c66822039d159da9ef98d92f676f9d. The branch file then holds those 40 characters and a newline: 41 bytes. Change one character of the file and every one of those three IDs changes.
This is a small teaching model. Use it to reason about a design, then check the details against your own system.
Build a real commit, byte by byte.
MockStruct / Quick Learn · 31 October 2026
https://mockstruct.com/learn/git-objects
Your one-minute recap
- Every object is named by the SHA-1 of its own type, length and contents.
- Identical content is stored once, so a branch copies nothing.
- A branch is one file holding a 40-character object ID and a newline.
Primary source: https://git-scm.com/book/en/v2/Git-Internals-Git-References
Built from Git: Pro Git — Git Internals: Git References (read 31 October 2026). Original MockStruct teaching model; no affiliation. Report a correction.