Skip to content

Latest commit

 

History

History
70 lines (53 loc) · 3.59 KB

README.md

File metadata and controls

70 lines (53 loc) · 3.59 KB
Raw

Disclaimer

Yes I reinvented something that was already there: git flow. Just use git flow ;)

gitnoob

gitnoob is just a small layer of abstraction for the most used git tasks. It helps adopting good practices and pushing clean code, while reducting the number of commands needed to push code. (Note : you can still use git commands directly if you want)

Using gitnoobs commands basically :

  • prevents you from coding directly into dev branch
  • prevents from merging breaking code into dev branch
  • force you to code inside feature branches
  • maintain a clean structure in dev and master

to get a list of available commands, use the -h argument.

installation

first you have to have ruby installed. Move the repository to a place you like, then you can run the install script (from the gitnoob folder) to add a symbolic link to the script in the /bin folder

chmod a+x install.sh
./install.sh

if you want helpfull short aliases to be added you can also execute the addaliases script (still from the gitnoob folder)

chmod a+x ./addaliases.sh
source addaliases.sh

prepare your repository

(here is a list of all the steps from nothing but you should pickup from where you actually are, you might already have a compatible repository)

  • build a new repository with git init
  • add your remote with git remote add origin https://github.com/user/repo.git
  • create the dev branch with git init git checkout -b dev
  • push all your branches with git push --all origin (if you already have a compatible architecture, make sure you are on the dev branch before starting using gitnoob)

Understanding gitnoob

Here is the simple concept of gitnoob workflow : The master branch hold your releases, the dev branch always contain a potentially releasable version of your code, and you are forced to create a feature branch if you wanna work.

the master branch

Each commit on this branch is considered a new viable version of your application, suitable for production testing and deployment. That branch cannot be accessed directly, to create à new version you use the -v argument form the dev branch.

the dev branch

The dev branch is the base where feature branches are created. You cannot commit directly inside dev but can begin a new feature or re-open an existing one with the -f option.

feature branches

Feature branches are the right place to code, "feature-" automatically prepend the git name of a feature branch. From here you can :

  • stage all you changes and commit them using -c [commit message]
  • update your feature with changes that happened in the dev branch with -u
  • go back to dev with -r
  • finish your feature and add it to dev with -p (as in "prune"), note that your feature will not be added to dev it tests fails (rails only for now)

Note : feature branches are automatically sync with origin (online) for collaboration purposes but you can branch locally from a feature branch if you want to, with the native git commands.

dicussion / improvements

  • todo: ask if feature branches should be pushed on creation and deleted on the remote on deletion (if present on remote)
  • todo: run system tests before creating a new version ?
  • todo: chose how to use remotes on features (if tracked maybe auto fetch and push)
  • add a basic init function that create the repository (done), add a remote and sets up all the needed branches to start a project fast
  • we might consider removing the dev branch as the code here is supposed to be ready to deploy. It looks like it could be simplified.