Skip to content

Latest commit

 

History

History
155 lines (106 loc) · 5.03 KB

AHitchhikersGuideToGit_Exodus.md

File metadata and controls

155 lines (106 loc) · 5.03 KB
Raw

Introduction

What is This?

This post is the second of a series of three blog posts that tries to provide a gentle introduction to git. It follows the premise:

If you know how a thing works, you can make it do what you want it to.

I'm assuming you either read the first part (TODO LINK), Genesis, or skipped it because you know all thats done in there.

Overview

Here, again, the overview, we're doing the second part today.

Genesis

Genesis (i.e. "Creation" or "Beginning"), covered some very basic things:

  • Configuring git so you can use it how you want. (Basic, aliases)
  • Creating your git repository (play god once!)
  • Creating your first git commits.
  • Learning what the repository is and where commits get stored.
  • Browsing through history.
  • Ignoring files.

Exodus

This tutorial, Exodus (i.e. "going out") covers:

  • Storing temporary changes without committing.
  • Navigating commits in git.
  • Sharing commits to a remote place.
    • Locally (huh, that's sharing?)
    • Via email
    • To a server
    • To a client
  • Working with others (i.e. non-linear).
    • Join things.
    • Linearize things.
  • Writing good commits.
  • Editing commits. (Actually not editing but it feels like that.)

Apocalypse

Apocalypse (i.e. "uncovering") will try to uncover some more advanced features of git, finally freeing your mind from your non-versioned filesystem:

  • Finding more information about code.
  • Finding causes of bugs in git.
  • Reverting commits.
  • Reviewing commits.
  • Travelling though time and changing history (you want me to believe you've never wanted to do that?)
  • Getting back lost things.
  • Let git do things automatically.

Disclaimer

I wrote this tutorial to the best of my knowledge and experience, if you spot an error or find something important is missing, be sure to drop me a message!

Preparation...

Don't forget your towel, whatever you do! If you still don't have one, go, get one! Now!

Branching

What Is a Branch?

You've waited long for this in the first tutorial, desperately - and I failed you. I'm sorry, wanted to cover the groundwork first. So let's get to this:

Do you recall that weird HEAD file in the .git directory I promised to explain a bit more last time? That HEAD file was a pointer to the current commit which contained the files we compare our filesystem with. Furthermore we saw that this HEAD file actually didn't reference the commit directly. It referenced some master file:

$ cat .git/HEAD 
ref: refs/heads/master

This file, refs/heads/master, contained just the ID of the current commit. This file is the master branch.

So to summarize: a branch is nothing more than a pointer (i.e. alias) to some commit.

How Can We Use Branches?

Let's try out the git branch command:

$ git branch
* master

The branch command shows us what branches are available. Obviously there is this master branch there - which corresponds finely to the refs/heads/master file.

Going Back in Time

So let's time travel a bit!

$ git log
commit 7b4977cdfb3f304feffa6fc22de1007dd2bebf26
Author: Lasse Schuirmann <[email protected]>
Date:   Fri Feb 27 16:39:11 2015 +0100

    README: Add usage instructions

commit ec6c903a0a18960cd73df18897e56738c4c6bb51
Author: Lasse Schuirmann <[email protected]>
Date:   Fri Feb 27 14:12:01 2015 +0100

    Add README

No let's do the magic:

$ git checkout ec6c903
Note: checking out 'ec6c903'.

You are in 'detached HEAD' state. You can look around, make experimental
changes and commit them, and you can discard any commits you make in this
state without impacting any branches by performing another checkout.

If you want to create a new branch to retain commits you create, you may
do so (now or later) by using -b with the checkout command again. Example:

  git checkout -b new_branch_name

HEAD is now at ec6c903... Add README
$ cat README
Hello World!
$ cat .git/HEAD
ec6c903a0a18960cd73df18897e56738c4c6bb51

So what we've done is actually resetting the files in our directory to the state of the commit ec6c903. In addition to that the HEAD file in our repository was set to that commit too so when we commit something, ec6c903 will be the parent of the new commit.

Committing on a Branch

When making a commit on a branch the current branch gets updated automatically, i.e. it points to the new commit.

TODO

Exodus (i.e. "going out") will cover:

Store temporary changes without committing.
Navigating commits in git.
Sharing commits to a remote place.
    Locally (huh, that's sharing?)
    Via email
    To a server
    To a client
Working with others (i.e. non-linear).
    Join things. git cat-file -p
    http://www.quora.com/How-does-Git-merge-work
    Linearize things.
Writing good commits.
Editing commits. (Actually not editing but it feels like that.)