This is source for a Ruby on Rails (Rails) web application that acts as a trial platform for preference aggregation using the Davenport algorithm for finding Kemeny orders.

The application was begun at the end of 2010, abandoned, and picked-up again at the end of May, 2019.

Development

• You will need mySQL.
  • On MacOS with Homebrew, brew install mysql
  • run it with /opt/homebrew/opt/mysql/bin/mysqld_safe --datadir\=/opt/homebrew/var/mysql
• You will need to have the Davenport library installed.
  • See https://github.com/wbreeze/davenport
• Check out the repository: git checkout [email protected]:wbreeze/socelect.git
• Change to the repository directory: cd socelect
• Install the Ruby programming environment at the level indicated in the file, .ruby-version. One way is to use the Ruby environment manager (rbenv).
• Run bundle install
• Start the Rails server for development: rails server --environment=development
• Visit the locally hosted development site to try the app locally

Delayed-job

For testing the queueing and delayed processing in development, use the command delayed_job --pool=result_state:1 --pool=result_compute:2 start. The Capistrano deploy for production is set-up to run the same.

The result_state queue should have only one worker, because it is used to serialize the state changes. There can be as many workers as you like on the result_compute queue.