A guide to git

Jacob Mattingley

Git is a powerful version control system. It's distributed, and has lots of good features. However, the interface developed over time, so it's a little bit weird to use. This is a short guide to git.

Throughout, we'll use kiwi as our client computer, and ant as our server. By no means is this the only way of using git; it's just a convenient way that works for me.

These are the available pages:

This page, which shows some useful features.
Creating and normal use of a git repository.
Adding server support.
Some configuration options.

Some useful features

Here's a list of commands which begin to explain why you'd bother using git:

git log: Show a list of previous commits. Note the long hex code — you can use that to refer to individual commits.
git grep string: Search all tracked files for string.
git diff: Show all changes since the previous commit.
git reset --hard: Blow away all changes since the last commit. (Use caution!)
git diff HEAD~10 file: Show changes to file since 10 commits ago.
git checkout file: Get the version of file that we last committed.
git checkout fa3be file: Get the version of file in commit fa3be. We only need enough characters in the commit id to make it unique.
git checkout fa3be file: Get the version of file in commit fa3be. We only need enough characters in the commit id to make it unique.
git show HEAD^:file: Show the second-to-last committed version of file.
git show fa3be:file: Show the version of file stored in commit fa3be.