Recording Changes

First-time setup

git config --global user.name "Your Name"
git config --global user.email user@domain.tld

Git is a distributed system, so it uses email addresses to uniquely identify users.

• try `git help config`

Creating a repository

git init repo
cd repo

This creates a new, empty repository which lives only on your computer (for now).

echo git > cool_stuff.txt
git status

• run `git init` without argument to start tracking existing files in `./`

Staging changes

git add cool_stuff.txt
git status

• you can also add entire directories with `git add`
• make changes in different parts of a file and try `git add -p`

Committing (to) a version

git commit
git status

• you can skip `git add` for files known to Git using `git commit -a`

Viewing history

---

gitk GUI separate tool tig ncurses separate tool git log text included in git

---

Every commit contains a snapshot of the entire repository; unlike checkins in RCS or Subversion, which operate on individual files or directories, respectively.

The why of commit messages

• the who/when/what of a change is recorded automatically
• commit messages primarily explain the "why"
• they summarize changes for history overviews (e.g. in gitk)
• they may contain pointers to ticket IDs for future reference

The how of commit messages

First line: very short summary, what changed?

After a blank line follows the answer to the "why". What was
the motivation for the change? How is the need addressed by
this commit?

If applicable: Which alternatives were considered, and why
were they discarded?

Closes #bugid.

Training Time

git config --global user.name "Your Name"
git config --global user.email user@domain.tld

1. create a new, empty repository
2. commit some changes
3. view the resulting history

• Try `git revert HEAD~1`. What does it do? Can you revert other changes?
• Look at the contents of the `.git` directory - what changes at each step?