Clowder uses OAuth for authentication, with the current provider being GitHub. At your local site, you need to create a GitHub OAuth application with your own Client ID and Client Secret. Set up environment variables containing these values, or put them in a .env file in your source directory:

echo "CLOWDER_GH_CLIENT_ID=aaaaaaaaaa" > .env
echo "CLOWDER_GH_CLIENT_SECRET=aaaaaaaaaa" >> .env

For the moment, we depend on crates that depend on Rust nightly (see SergioBenitez/Rocket#19 for our main blocker). You will likely want to use Rustup to install the nightly version of Rust.

Clowder requires a database to be created that is accessible to the user running the service. Using Postgres (at least on FreeBSD), this looks like:

# service postgresql initdb
# service postgresql start
# su - postgres
$ createuser ${username}    # with a username like, e.g., clowder
$ psql postgres
postgres=# create database clowder;
postgres=# grant all on database clowder to ${username};

You should set the database URL in an environment variable, or in a .env file within your source directory:

$ echo "export DATABASE_URL=postgres://localhost/clowder" >> .env

Once the database has been created, we use the Diesel ORM to initialize it:

$ cargo install diesel_cli
$ cd path/to/clowder/source
$ diesel migration run

Once Rust and the Clowder database have been set up, you can build and run Clowder!

$ cargo build
$ cargo run

You can set CLOWDER_FAKE_GITHUB_USERNAME in your environment (or .env file) to provide a username to be treated as if it were obtained from the normal GitHub OAuth workflow.

For development purposes, I like to use cargo-watch to rebuild whenever I change a source file:

$ cargo watch --ignore '*.swp' --exec run

This goes quite nicely with LiveReload.