You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Here is where a brief description of the project should go.
It may be worthwhile to list major project contributors here, especially if this is a closed source project as one might need to reach out for questions later on.
Setting Up
Explain any specific project dependancies here.
Deploying
Explain how to deploy this project. Maybe minimum computer specifications or browser requirements are listed here as well.
Version History
0.0.1
This is an update to the version history.
Version history should be thoroughly updated and correspond to project tags.
I keep meeting people who struggle to wrap their heads around the proper way to use Git + Github. This snippet is designed to explain Vincent Driessen's git branching model, at least as well as I understand it. Speacial thanks to Stephen Koch for being the true master here.
This tutorial is for Linux or OSX. Feel free to use Cygwin or write a fork for windows.
A way to think about Git and Github.
Milestones of milestones of milestones. In other words:
Open up a text editor.
Type "Hello World".
Save this file.
You have now created a "milestone" on your hard drive of this text.
You can now retreive that milestone by double clicking it to re-open it in your text editor.
This should be a concept you already understand quite well.
Change the contents of that file again. Add in your own text. Save it again.
By saving it again you've overwritten the previous milestone.
You can certainly redo the work (e.g. replacing all the text with "Hello World" and saving again) but the original work is gone otherwise.
Git saves milestones of milestones.
git commit -am "By typing this command I am saving a collection of saved files."
This is great because now we can roll back to old versions of files without having to retype. Aka "source control".
However, wouldn't it be great if we could further save milestones in the cloud?
Aka milestones of milestones of milestones.
Github -> Git -> Save
Github is two things:
git, in the cloud
a social network around source code
All you need to do to push to Github:
git push origin master
Now one could "clone" that repository on another computer and not just get the latest code but the complete revision history on another computer.
Setting up
Assuming your project is in a folder named "Project" on your Desktop.
Starting from scratch
cd ~/Desktop/Project
git init
git checkout -b develop
touch README.md
Open the README.md file you just created in your text editor. Describe your project. I've provided a basic template below for what it's worth. Save it.
Go to Github (or Bitbucket or whereever you want to save your code in the cloud). Create a new project.
If you're on Github, do not check Initialize this project with a README since you just made one.
Determine your SSH clone url. On Github it's probably something like [email protected]:USERNAME/PROJECT.git. Should be on the project's page somewhere.
Add your remote.
git remote add origin {{the link you just copied}}
Breaking that down
git :: The git command
remote add :: We're adding a remote connection for this repository
origin :: We're naming the remote "origin". You can also call this "github" or "bananasauraus" if you'd like.
Cloning an existing repository.
Determine your SSH clone url. On Github it's probably something like [email protected]:USERNAME/PROJECT.git. Should be on the project's page somewhere.
cd ~/Desktop
git clone {{the link you just copied}} Project
This creates a directory named "Project", clones the repository there and adds a remote named "origin" back to the source.
cd Project
git checkout develop
If that last command fails
git checkout -b develop
Updating/The Development Cycle
You now have a git repository, likely with two branches: master and develop. Now bake these laws into your mind and process:
####You will never commit to master directly.
####You will never commit to develop directly.
Instead, you will create feature branches on your machine that exist for the purpose of solving singular issues. You will always base your features off the develop branch.
This last command creates a new branch named "my-feature-branch" based off of develop. You can name that branch whatever you like. You should not have to push it to Github unless you intend to work on multiple machines on that feature.
Make changes.
git add .
git commit -am "I have made some changes."
This adds any new files to be tracked and makes a commit. Now let's add them to develop.