bk99.de entertain the web since 1997

bk99.de
HN · GitCurated

Understanding Git from its objects outwards

Summary

Mary Rose Cook explains Git by building up blobs, trees, commits, branches, merges and rebases step by step. Git stores objects in the .git/objects directory. Object names are derived from hash values.

Ideas

  • Blobs store immutable file contents.
  • Trees map names to stored objects.
  • Commits link a tree with parents and metadata.
  • Branches are mutable pointers to commit objects.

Insights

  • Git's interface becomes understandable once its data model is visible.
  • Content addressing simplifies verification, deduplication and exchange.
  • History forms a graph rather than a linear list.

Facts

  • Merge commits have several parents.

Recommendations

  • Examine a test repository with git cat-file.
  • Draw commit graphs before complex history operations.

References

Read the original article

Search the Web Archive

HN · 2009

Understanding Git as a graph of immutable objects

Charles Duan explains blobs, trees, commits and references as the connected data model behind Git's everyday commands. Git uses four central object types. Object names are derived from cryptographic hash values.

HN · 2009

Pro Git explains version control from the inside

Scott Chacon's freely available book explains Git from the everyday workflow to the object model and distributed server operation. Pro Git was originally published in 2009. The book is licensed under a Creative Commons licence.

HN · 2009

Deriving Git from simple file system ideas

Tom Preston-Werner develops Git's basic model step by step from files, snapshots, hash values, pointers and shared history. Git uses content-addressed objects. Commits form a directed graph through parent references.