• Commit -- A point in time snapshot of your repo • Reference -- Something that makes it so you don't have to remember a 40 hexadecimal digit sha1 hash • Branch -- A pointer to the head of a group of commits • Remote -- A thing you push to and pull from • Pull Request -- How you get your code reviewed (unless you're on kernel) • Github -- Magic
4. Edit some files 5. Commit 6. Repeat 4 and 5 until feature/bug is complete 7. REBASE to create a history that I won't mind reading through 8. Pull any changes that have happened on non-forked master in the interim and rebase on top of those if you want a linear history otherwise do nothing for this step 9. Open pull request from your branch on your fork into the canonical repo's master branch 10. Profit
~ selects merge parents (HEAD~, HEAD~2) • Mix and match (don't actually though) • HEAD is hard coded and always points to the current head of whatever branch you have checked out • Almost everything can be treated as a ref
(a tree where the leaf nodes are blobs) • Commit (a pointer to: a tree, n parent commits) • Tag (a pointer to a commit) • (IGNORE THIS) something something packfiles