Git from the Bottom Up – Blobs and Trees

Blobs

Blobs are the fundamental data units in Git, representing file contents without any metadata. Alan Underwood explains that blobs are identified by a SHA1 hash of their size and contents, ensuring consistency across systems 1. Michael Outlaw adds that blobs are stored in trees, which maintain the file structure within a repository 2. This design allows Git to efficiently manage data by using the same blob for identical content, regardless of file names or locations 3.

   

Trees

Git uses trees to organize blobs and track file system structures. Joe Zack notes that trees are created by writing the contents of the index into a tree object, which is then used in commits 4. Alan Underwood highlights that the git ls-tree command shows the tree structure at different levels, depending on where it's run in the directory 5. This contextual view allows users to see how blobs and trees are interconnected within the repository 6.

   

Commits

Commits in Git function as snapshots, linking trees and blobs while incorporating metadata. Alan Underwood explains that commits are unique due to their inclusion of author names and timestamps, which affect their SHA1 hashes 7. Michael Outlaw discusses orphaned commits, which occur when a commit isn't referenced by any branch, highlighting Git's garbage collection process 8. Joe Zack emphasizes the importance of metadata in understanding the structure and history of a repository, noting that commits point to trees, which in turn organize blobs 9.