What process do you use for WordPress development? [closed]

For the record, I mainly make whole Web sites and plugins, and deploy them. My workflow is very Ruby- and git-heavy.

To start out on a new project, I have a shell script which takes care of the whole business of setting up a new vhost and checking out the latest tag of WordPress (from our own git repository, which tracks svn).

The basic shape of a whole Web site is a git repsotory at wp-content. That contains a Capfile (capistrano’s Makefile eqiuivalent) and a YAML configuration file which together take care of deployment ( http://github.com/dxw/wp-capistrano ). Also inside that repository I add the theme and plugins as git sub-modules (yes, we maintain git repositories for third party plugins too – we like to use the latest version that we’ve personally tested).

For the theme, I have a code-generation tool/framework ( github.com/dxw/wp-generate ). It means less thinking about where code should go, and it has a natural method of seperation between the View and the Model/Controller.

When writing plugins I use cucumber/webrat to do test-driven development ( github.com/dxw/cucumber-wordpress ).

And for migrating development databases to production, it’s usually just a case of copying the dump over (WP_SITEURL and WP_HOME are set by capistrano on the staging/production machines so no search/replace).

I can’t imagine how many hours I’ve saved with these scripts.

Leave a Comment