What is Git?

Baby don't hurt me / don't hurt me / no more

Transcript

1. What is Git
   Baby don’t hurt me / don’t hurt me / no more
   

   View Slide

2. Page
   So, what is Git?
   › Git is an application that provides Distributed Source Control.
   › It provides features similar to TFS in that it allows us to version
   control our code, tag these versions with metadata, stash,
   branch, merge, and a bunch of other bits and bobs.
   › The windows version is called msysgit
   (https://msysgit.github.io/)
   › Git can be used from either a command line interface (CLI), or
   from a few GUI-driven applications, such as GitExtensions
   (https://code.google.com/p/gitextensions/)
   / Copyright ©2014 by Readify Pty Ltd
   2
   

   View Slide

3. Page
   Distributed Source Control
   › TFS is centralized, Git is distributed.
   › With TFS, when you create a workspace mapping from a repository to
   your PC and retrieve the code, you are retrieving a snapshot of the latest
   version of the code from the TFS server.
   › With Git, when you Clone a repository, you retrieve the entire repository
   from the nominated Origin, including all history.
   › This means with git that once you Clone a repository, you have a complete
   working backup of the Origin.
   › Once you have Clonedyou can then work completely offline with the
   repository until you decide you want to sync with Origin.
   / Copyright ©2014 by Readify Pty Ltd
   3
   

   View Slide

4. Page
   Local vs Remote
   › Work locally, synchronize remotely.
   › Local commands: add, commit, branch.
   › These commands let you modify your local git
   repository while you work.
   › Remote commands: pull, fetch, push.
   › These commands let you interact with your remote
   repository and synchronize your work.
   / Copyright ©2014 by Readify Pty Ltd
   4
   

   View Slide

5. Page
   Working locally
   / Copyright ©2014 by Readify Pty Ltd
   5
   When working locally, files are in
   one of three states: committed,
   modified, or staged.
   Committed files are stored safely
   in your local .git database.
   Modified files are files in your
   working directory you have
   changed but not yet staged or
   committed.
   Staged files are modified files
   that have been marked to be
   added to your next commit.
   

   View Slide

6. Page
   Staging visualised
   / Copyright ©2014 by Readify Pty Ltd
   6
   

   View Slide

7. Page
   How do I use it?
   › Lets assume we have a local repository cloned and we are ready to write some code.
   › First, create a branch with git checkout –b “branchname”. This will keep our changes
   isolated from Master until they have been reviewed in a PR.
   › Now code away!
   › When you have satisfied your need to code, you can see what changes you have made
   with git statusor git log. Or just jump into GitExt
   › When you are ready to ‘check in’ your code, you can use git add –Ato “stage” all of your
   changes.
   › Once you are happy with your staged changes, commit them locally with git commit –m
   “Commit message”
   / Copyright ©2014 by Readify Pty Ltd
   7
   

   View Slide

8. Page
   Push, pull, fetch!
   › Once you have been working for a little while you may want to integrate
   any changes in the master branch with your work.
   › You have a couple of options! Pull, or fetch.
   › When you pull, changes from the remote branch you pull from (Master in
   this case) will be merged directly into your local workspace, and you may
   have to resolve changes.
   › When you fetch, you pull the changes down into your git repository, but
   the working area is not updated. You can then then optionally merge
   those changes into your branch if you wish.
   › Once you have finished merging, Push your changes up!
   / Copyright ©2014 by Readify Pty Ltd
   8
   

   View Slide

9. Page
   Under the hood
   › Git works in such a way that you only ever have a single copy of the code in your working
   directory.
   › Commits store a snapshot (not a diff) of the file changes you have made in the git
   database, along with a reference to the previous commit.
   › Commits therefore end up forming a “directed acyclic graph”
   › When you check out a commit your working directory is rebuiltbased on the tree of
   commits from the one you have checked out.
   › This means when we branch, we still only have a single copy of the source locally, we
   don’t need separate folders for separate branches.
   › When we check out a different branch, git will rebuild our working directory based on the
   tree of commits cascading down from that branch.
   / Copyright ©2014 by Readify Pty Ltd
   9
   

   View Slide

10. Page
    Branching
    › Branching in Git is a lightweight operation, as branches are just pointers to
    a commit.
    › A branch therefore is just a reference to a commit.
    › Branches come in a couple of flavours: local and remote. Local branches
    are those on our machine that point to commits on our local repository.
    › Remote branches are of the form origin/{branch}, and they let us track the
    state of our remote repo. Git automatically updates them every time we
    communicate with our remote for some reason.
    / Copyright ©2014 by Readify Pty Ltd
    10
    

    View Slide

11. Page
    Branching workflow
    › With Git you are always working in a branch.They are non-optional.
    › The central branch or ‘trunk’ is referred to as master. Masteris the default
    branch created when the repository is initialized. When working you
    branchfrom master, and when done you mergeback into master.
    / Copyright ©2014 by Readify Pty Ltd
    11
    

    View Slide

12. Page
    Glossary
    › Clone: akin to establishing a workspace mapping and pulling down the code in TFS, cloneretrieves an entire
    repository for us.
    › Pull: akin to “Get Latest” in TFS, pull retrieves the latest changesetsfrom our upstream Remote repository and
    attempts to merge them in with our current work.
    › Fetch: similar to pull, this retrieves the latest changesetsfrom our upstream Remote repository, but does not
    attempt to merge them in with our current work.
    › Add: add marks files that you have changed as ready to commit. Only the files marked by ‘add’ will then be
    committed on your next commit. Other modified files will not be committed.
    › Commit: saves the work you have staged into the git repository.
    › Push: akin to ‘check in’ in TFS, this pushes any changes you have made locally up to the remote repository. If it
    has changes that you haven’t yet pulled down and integrated, this will fail.
    / Copyright ©2014 by Readify Pty Ltd
    12
    

    View Slide

13. Page
    Further learnings
    › http://think-like-a-git.net/ (How git uses graphs to
    track changes, and how we can use them to make our
    lives easier)
    › http://rogerdudler.github.io/git-guide/ (A nice brief
    guide to working with git)
    › http://jake.ginnivan.net/explain-git-with-d3/ (A
    sandbox for playing with git concepts)
    / Copyright ©2014 by Readify Pty Ltd
    13
    

    View Slide