Git Workflow

Overview

Git Server set up on dbarchowsky.com.

Creating a new repo out of an existing site

• Navigate to the site's root directory.
• Create a Git repository.

$ git init

• Make a .gitignore file (images, wordpress install files, etc.)
• Set username and email (necessary for each repo)

$ git config --global user.name "Damien Barchowsky"
$ git config --global user.email "dbarchowsky@gmail.com"

• Add existing files to Git repo, confirm the list of files, and commit the changes.

$ git add .
$ git status -s
$ git commit -m "Initial version"

Working with existing projects

• Make sure the 'dev' drive is available on the current client:

subst m: "\\littledamienii\develop"

Using 'Git Bash' navigate to the project directory.

$ cd /m/littledamien/littledamien_web

• Make a new Git branch

$ git branch -b mynewbranch

• Edit, add, and commit as usual
• When the update is complete, merge the branch with 'master'
• Then move the commits to the main repo:

$ git push ssh://[uname]@[domain].com:[port]/path/to/repo master

Configuration

• Physical origin repos located beneath d:\shared\git
• To make a new project:

d:
cd \shared\git
git clone ..\..\develop\path\to\project_root
cd .\path\to\project_root
git init 
rem <<< is that last 'git init' step necessary ??? >>>

• At least for now, when fetching & pulling vial ssh, the root directory is the Git install directory (C:\Git).

• A symbolic link has been created between C:\Git directory and the Git base directory (D:\shared\git)

cd \Git
mklink /d "refs" "d:\shared\git"
rem symbolic link created for refs <<===>> d:\shared\git

• The symbolic link allows access to the git repo via

git clone|push|pull|etc. ssh://username@server/refs/path/to/myrepo