Mastering Git’s index
summary
The index is a path-to-blob mapping that doubles as a stat cache; commands shown manipulate it directly: using read-tree to populate an empty index or import a subdirectory (with prefix) to create a new repo subtree, switching indexes via GIT_INDEX_FILE, and using plumbing like update-index and checkout-index. Using --index-filter inside filter-branch avoids checkouts and runs roughly 10–20× faster. Advice covers skip-worktree versus assume-unchanged, sparse-checkout to manage large working trees, and running git update-index --refresh to avoid costly git status scans.
AI-written summary · about
chapters (22)
- 0:00Intro
- 0:25What is Git
- 2:40Git attributes
- 3:23Gateless
- 4:42Out of interest
- 9:03The index
- 11:27History of Git
- 13:30Index
- 15:22Reset
- 23:39Why use Git
- 27:00Clearing the index
- 29:18GetRetreive
- 30:54GetPrefix
- 34:20Mono Repo
- 37:24Assume unchanged
- 39:40Skip work tree
- 43:17Path check out
- 48:26What is in the set cache
- 52:39Get update index refresh
- 53:38Git diff
- 55:31Storytime
- 59:21Performance
