Knowledge is Power: Getting out of trouble by understanding Git
summary
Explains Git’s low-level model (blobs, trees, commits) and shows practical recoveries: how refs are just pointers so branches are cheap, how git reflog records every ref move (giving a two‑week grace before GC), how git rebase --exec runs tests during a replay to stop on failures, and how git bisect automates a binary search for regressions. Demonstrates a “hidden” object trick using git hash-object then reading it with git cat-file, and shows using reset and reflog to restore lost commits.
AI-written summary · about
chapters (22)
- 0:00Intro
- 5:37The blob
- 5:58Types of blobs
- 6:26Getting an empty Git repository
- 7:13Somefile
- 12:07echo
- 12:34file systems
- 14:35Refs
- 16:14Branching
- 17:39Reset
- 18:42Ref Log
- 19:55Garbage Collection
- 21:07Save Space
- 21:49Pack Files
- 23:48merges
- 25:53fastforward merge
- 27:49merge strategies
- 31:28rebase
- 41:01hash objects
- 42:57getcatfile
- 43:53git repository
- 46:27recap
related talks
56mKnowledge is Power: Getting out of Trouble by Understanding Git
39mGetting out of trouble by understanding Git
48mBe the .txt you seek: Getting out of trouble by understanding Git
25mUnderstanding git - even the scary parts
32mGit, back to the future
35mThe Git you don't know
28mHow to Really Get Git
41mKeynote: Wrapping Your Head Around Git