Why to choose Git Repository

There are numerous internet articles which will tell you difference between Git and other popular repository system like subversion, CVS etc. But I didn’t want to discuss comparison rather I will let you know what the benefits Git can delivers to us, so that we can clearly make our mind whether to go for Git. Below are some of benefits

• No Internet required to work on, you can work offline.
• Branching is a light weight process
• You can stack your uncommitted changes and can popup them once you are ready.
• More flexible workflow
• Data redundancy & replication
• Works well with open source project or projects where contributors may take months to contribute there code.

Let’s review each of them in detail

No Internet required to work on, you can work offline – First steps to use Git is to create a local repo at your machine. You can use Git Clone command for the same. Since you have a complete repo at local machine, we did not have to bother about internet connection. We can access complete history or revision and commit our changes. Once we get online and are ready to submit code then we can push code to remote server. Here I want to brief you with terminology Push. In Git operation commit is used to commit changes in Local repository and when you want to commit your changes into the remote repository from where you have cloned your Local repository you will be ‘Pushing’ your changes. The command Push can be used for the same. Similarly for getting changes from remote repository after clone operation you can use Pull operation.

Branching is light weight process – We all know advantages of branches. In Git creation of a branch is simply creating a Pointer. Git branching is so cheap that you can create hundreds of branch in a day and can delete them. It’s always good to work with branches so that main trunk will always be cleaner. In Git we can merge even two branches which don’t have common ancestors. We will look more on it in Branching and merging section.

Stash – In order to understand Git Stash feature lets consider a scenario which we all had encountered some or other way. Let’s consider that I am working on an issue1 and I am in middle of my changes. Suddenly my manager comes in and asks me to quickly fix a production bug. Unfortunately the changes required to fix production bug is in same file on which I had changes for issue1. Since I am not done completely with issue1 coding, I can’t check In my code. But I want to fix production issue which should be on latest code. One way is to backup file at our file system and replaces workspace file with latest from repository and later on when we done with Production changes we can copy backup file in to workspace.

Git provides us one easy way to do this as Stash. You can stash your uncommitted changes in Git and later on can unstash your changes to continue your work. In data source terminology you can push your uncommitted changes in to a Stack using stash command and later on can popup your changes. With stash command, Git will save your changes and will replace workspace file with latest from repository. Later on when you done with your critical issue fixing you can unstack your changes using apply option of stash command.

More Flexible Workflow (Stage/Index) – Working with repository we usually have below workflow

1.       Checkout/Update files from repository

2.       Modify files for relevant changes

3.       Commit file back to repository

In Git one more level/area is introduce before committing your changes. This area is known as Stage/Index. You can consider this level as an area where you are preparing your commit. Once you prepare your commit you can move ahead for commit. Once you are done with your changes then you need to add modified files to staging using add command.

$git add <filename>

Adding a file in staging is optional, You can use commit command with –a option to add a file into staging and then committing. The command commit with –a option will merge add and commit command.

Let’s consider same scenario as above, we are working on an issue1 and we need to pause our work to fix a production issue. This time changes required to fix production issue is a subset of changes done for issue1. In other words we can say that some of changes implemented for issue1 will be fixed production issue. But committing a file means committing all changes/modifications in a file. We can’t able to select a set of changes within a file to commit.  

Git enable us to do it using stage area. You can choose selected changes within a file and can add them to staging area. Only changes added to staging area will be getting committed. We can use Git add command in interactive mode.

Please refer link for more details http://git-scm.com/book/en/Git-Tools-Interactive-Staging

Git provides easy way to ignore files from tracking. It provides us fault tolerance as data redundant on multiple machines. Other than above benefits, Git provide lot more benefits which we can explore more when we actually start using Git.

Conclusion: Overall you will be benefited by using Git. It provides lot more flexibility which I have not able to cover above. There are lots of client tools available for Git to use like Tortiose Git, SmartGit etc. For eclipse lovers there is a Egit plugin to make life easier. There is only one disadvantage which I see with Git is about space utilization. Since a complete repository is replicated on each developer’s machine it will definitely not beneficial or effective in case you have limited space. But in my opinion Data storage are so cheap now a days that it can be easily overcome.