Understanding

the Workflow of
Version Control
›
presented by TOWER Version control with Git - made easy

The Basics
1

Work on an Existing Project

Start a New Project
$ git clone <remote-url>
$ git init
The “git clone” command is used to
Executing the “git init” command in the root download a copy of an existing repository
folder of your new project creates a new and from a remote server. When this is done, you
empty Git repository. You’re ready to start have a full-featured version of the project on
getting your files under version control! your local computer – including its complete
history of changes.

2
File Status
Files that aren’t yet under version
Work on Your Files
?
control are called “untracked”…
Modify, rename and delete files or add new …while files that your version
ones. Do all of this in your favorite editor / control system already knows
IDE / file browser – there‘s nothing to watch about are “tracked” files.
out for in this step! A tracked file can either be
“unmodified” (meaning it wasn’t
changed since the last commit)...

m
...or “modified” (meaning it has
local changes since it was last
committed).

3
$ git status
Keep the Overview #
# Changes not staged for commit:
$ git status # modified: about.html
# deleted: robots.txt
The “git status” command tells you what #
happened since the last commit: which files # Untracked files:
did you change? Did you create any new ones # login.html
or delete old ones? #
no changes added to commit

Add Files to the “Staging Area” $ git add about.html

#
$ git add <filename> # Changes to be committed:
# modified: about.html
Only because a file was changed doesn’t #
mean it will be part of the next commit! # Changes not staged for commit:
Instead, you have to explicitly decide which # deleted: robots.txt
changes you want to include. To do this, you #
add them to the so-called “Staging Area” # Untracked files:
with the “git add” command. # login.html

Commit all Staged Changes

$ git commit -m "message"
$ git commit -m "Updated about page"
A commit wraps up all the changes you
previously staged with the “git add” [master 9d3f32b] Updated about page
command. To record this set of changes in 1 file changed, 29 insertions(+)
Git’s database, you execute the “git commit”
command with a short and informative
message.

Keep the Overview

$ git status
$ git status #
# Changes not staged for commit:
Running the “git status” command right after
# deleted: robots.txt
a commit proves to you: only the changes
#
that you added to the Staging Area were # Untracked files:
committed. # login.html
All other changes have been left as local #
changes: you can continue to work with them no changes added to commit
and commit or discard them later.

Inspect the Commit History

$ git log
$ git log
commit 9d3f32ba002110ee0022fe6d2c5308
The “git log” command lists all the commits Author: Tobias Günther <tg@fournova.c
that were saved in chronological order. This Date: Mon Jul 8 09:56:33 2013 +0200
allows you to see which changes were made
in detail and helps you comprehend how the Updated about page
project evolved.

Branching & Merging

1

Understanding Branches
BUGFIX #32 C4

FEATURE A C1 C3 C5 C7

FEATURE B C2 C6

Start a New Feature We often have to work on multiple things

in parallel: feature X, bugfix #32, feature Y…
$ git branch <new-branch-name> This makes it all too easy to lose track of
where each change belongs. Therefore, it’s
Whenever you start a new feature, a new
essential to keep these contexts separate
experiment or a new bugfix, you should
from each other.
create a new branch. In Git, this is extremely
fast and easy: just call “git branch <new- Grouping related changes in their own
branch-name>” and you have a new, context has multiple benefits: your
separate context. coworkers can better understand what
happened because they only have to look at
Don’t be shy about creating new branches: it
code that really concerns them. And you can
costs you nothing.
stay relaxed, because when you mess up,
you mess up only this context.
Branches do just this: they provide a context
that keeps your work and your changes
separate from any other context.

2 HEAD Branch
C6 C7 feature-b

Switch Contexts C1 C4 C5 master

$ git checkout <new-branch-name> C2 C3 feature-a HEAD

To start working on a different context, you

At each point in time, you can only work in
need to tell Git that you want to switch to it.
one context – the context of the currently
You do this by “checking out” the branch with
checked out branch (which is also called the
the “git checkout” command.
“HEAD” branch in Git).
Every commit you make – until you switch
Your project’s working directory contains the
branches again – will be recorded in this
files that correspond to this branch. When
branch and kept separate from your other
you check out a different branch (make
contexts.
it “HEAD”), Git replaces the files in your
working directory with the ones that match
this branch.

Integrate Changes
$ git merge <branch-to-integrate>

When your new feature is ready, you might

want to integrate it into another branch (e.g.
your production or testing branch).
First, switch to the branch that is supposed
to receive these changes. Then, call the “git
merge” command with the name of the
branch you want to integrate.

Sharing Work via

Remote Repositories
1

Track a Remote Branch

Publish a Local Branch
$ git checkout --track
<remote/branch> $ git push -u <remote>
<local-branch>
If there’s an interesting remote branch that
you want to work on, you can easily get To share one of your local branches with
your own local copy. Use the “git checkout” your teammates, you need to publish it on a
command and tell it which remote branch remote server with the “git push” command.
you want your new local branch to base off.

Local & Remote Repositories

2 MODIFY, ADD &
DELETE FILES
MAKE COMMITS

LOCAL REMOTE
Stay Up-To-Date
SHARE WORK
COMPUTER SERVER

About Remote Changes LOCAL

REPOSITORY COLLABORATE
REMOTE
REPOSITORY

VIEW HISTORY
$ git fetch <remote> ADD & DELETE
BRANCHES

When collaborating with others on a

project, you’ll want to stay informed about As Git is a so-called “decentralized” version
their changes. The “git fetch” command control system, a remote repository is
downloads new changes from a remote optional. In fact, everything we did until now
repository – but doesn’t integrate them into happened on your local machine, in your local
your local working copy. It only informs you repository – no internet/network connection
about what happened on the remote, leaving was necessary.
the decision on what to integrate to you. However, if you want to collaborate with
others, you need a remote repository on a
server. You don’t have to share all of your
work though: you can decide for each of your
local branches if you want to share it or not.

Integrate Remote Changes

$ git pull

To integrate new changes from the remote

repository, you simply call “git pull”.
This will update your current HEAD branch
with new data from its counterpart branch
on the remote. The changes will be directly
merged into your local working copy.

Upload Local Changes

to the Remote Server
$ git push

To upload the local changes you made in your Version control with Git - made easy

current HEAD branch, all you have to do is 30-day free trial available at
call “git push”. www.git-tower.com