Contained in this document are instructions for how to get the ecore development environment to work. Note that I'm developing primarily on a Mac, and while the instructions are suitably generic and will probably work on Linux, adjustments might need to be made to get this to work on Windows.

Step 1. Download VirtualBox
VirtualBox is available from here. You'll need it, as it runs the underlying VM that we'll be working off of.

Step 2. Make sure that you have a ruby binary installed.

Step 3. Install Vagrant
Vagrant is our virtual machine infrastructure glue, and it's written in ruby. It will run the chef recipe needed to get everything working once you download everything. Grab it from the downloads page and install it. Your package management system might also have it, but note that as of 2013-04, the rubygems version is out of date.

Step 4. Sync the code from github
We're on github now, more as this continues. If you don't have a git binary in your path, you'll need one from ports or apt or wherever you get your software. http://book.git-scm.com/2_installing_git.html has instructions for how do do this. You'll need to set up your git environment some, like setting your username and email if you'll want to commit. http://help.github.com/set-your-user-name-email-and-github-token/ has how to do that. Lastly, with how to fork the repo to make commits, check out: http://help.github.com/fork-a-repo/. Anyone who had commit access to the repository before, I'd be happy to restore that access, or you can send me a git pull request.

Step 5. Boot up vagrant
I'll write up a vagrant primer in a bit, but here are the basics.
  • vagrant up - Starts your VM up, creating it from scratch if needed. Note that if you have never done this before, it has to download the base box from bonci.net, so please don't go getting rid of the box entirely. Note that it is still available if you do a vagrant destroy. You need to specifically trash the underlying box.
  • vagrant halt - Preserves the vagrant state and shuts the machine down gracefully
  • vagrant destroy - Trashes the virtual machine. Don't worry, vagrant up will recreate you a clean environment, but there's no way to get your data back.
  • vagrant ssh - ssh's into the machine as root.
All commands must be done from the everything2/vagrant folder. For instance to start it the first time:
cd everything2/vagrant
vagrant up


Note that it will take a little while to get it up and running the first time, as it is installing packages and configuring the system.

Lastly, in your VM host OS, go to http://localhost:8888. This will bring you to the frontpage of the ecore setup. All of the passwords have been set to 'blah' for now, so you can log in as root and away you go. Note that as of the time of these instructions, there are still a few rough edges which need to be smoothed, but we'll get there.

Final notes:
  • This is running the same version of perl and other software as production, so we should be seeing the same warnings
  • Stored procedures aren't captured in the node dump right, so they aren't imported properly. More on that later.
  • The root key to the ssh machine is in the vagrant subdirectory. It doesn't have a password
  • The root password, if you need it, is 'vagrant'