What is this thing?
The Food Rescue Robot is a Rails 3 web application for managing the logistics of just-in-time (i.e., warehouse-less) food rescue. It keeps track of donors, recipients, a pickup and delivery schedule, and all the volunteers responsible for doing the work. It sends reminders to users about their pickups. The site also logs how much food is rescued, and can automatically issue receipts for donations.
Who uses it? (And how can I?)
Currently, the "Food Rescue Robot" is live at http://robot.boulderfoodrescue.org and is used by a number of food rescue organizations cities around the world. If you're keen to use it, shoot an email to caleb at boulder food rescue (dot) org, and I can set you up with an account. Alternatively, if you would like to fork the source and hack on it yourself you are welcome to. This code is licensed under a Beerware-style license:
As long as you retain this notice you can do whatever you want with this stuff. If we meet some day, and you think this stuff is worth it, you can buy me a beer in return.
Who is responsible?
Most of the code was written by Caleb Phillips, a co-founder of Boulder Food Rescue (http://www.boulderfoodrescue.org) and adjunct Computer Science professor at the University of Colorado, Boulder. There is a fork maintained in Boston by Rahul Bhargava (MIT Media Labs), and some early design work was done by University of Colorado interns Sean Weise and Zac Doyle under awards from the Casey Feldman Foundation. As of May 2016, Rylan Bowers (http://rylanbowers.com/) is leading development on the master branch and we are growing a small team of hackers to maintain and improve the codebase.
How can I help?
If you want to help with development, feel free to fork the project. If you have something to submit upstream, send a pull request from your fork. If you're trying to setup a dev environment, keep reading.
Alternatively, if your OSS time is limited, you can donate to Boulder Food Rescue via several options. Please include a note that indicates you are donating for Robot Development in the notes.
General Tech Overview
The software's main goal is to keep a record of what food got picked up, when, and by whom and, to some degree, where it was delivered. The key database relations are:
- Logs - history of what was picked up, when, how and by whom. Log entries are generated from the schedule three days in advance.
- Schedule Chains - Schedule of each route, where food is taken from one or more donors and given to one or more recipients
- Locations - Donors or recpients, places where food comes from or goes to
- Volunteers - The main user table, which authentication happens against
- Regions - Keeps, e.g., Denver's stuff separate from Seattle's stuff
Besides that, there are some secondary relationships:
- Food Types - e.g., produce or bakery
- Transport Types - e.g., bike or car
- Scale Types - e.g., floor scale or bathroom scale
- Absences - tracks when volunteers leave
- Assignments - list of which region each volunteer belongs to
- Cell Carriers - hacky way to do email->sms that should be replaced
There is a basic JSON API, which is provided by some controller methods (look for
responds_to :json). You
can see the routes with
Preparing a Development Environment
- Ruby 2.3.7
- Postgresql 9.3 or greater (runs on at least 9.4.4)
- A reasonable operating system (e.g., Linux or Mac, Windows if you're saucy)
- Various dependencies for the above
Clone this repository:
git clone https://github.com/boulder-food-rescue/food-rescue-robot.git
cd into the directory:
Linux or Mac
Set up your environment:
bundle exec bin/setup
This will check your system for prerequisites (e.g. Ruby version), install dependencies, create a
.env file, create any missing databases and database users, load the database schema and seed development data.
You may need to insert the secret key manually. Check the
.env file's SECRET_KEY_BASE. If absent, generate a secret key by using the command below:
bundle exec rake secret
Manual or Windows
.env, and replace
SECRET_KEY_BASE with the value of:
Create a development and test Postgres database and user:
CREATE ROLE bfr_webapp_db WITH LOGIN; CREATE ROLE bfr_webapp_db_test WITH LOGIN; CREATE DATABASE bfr_webapp_db OWNER bfr_webapp_db; CREATE DATABASE bfr_webapp_db_test OWNER bfr_webapp_db_test;
Load the datbase
To start a server, run
bundle exec rails server
Also, beware that some crucial functions, like generating log entries from the schedule, and sending emails are executed by cron, daily or weekly. If you want to work on those bits, you may be keen to run these:
bundle exec rake foodrobot:generate_logs bundle exec rake foodrobot:send_reminders bundle exec rake foodrobot:send_weekly_summary
Generating Sample Data
The seeds command generates a regular volunteer and an admin volunteer for you. Please review seeds.rb. You can make more regions / volunteers with this code:
$ rails console region = Region.create(name: "Boulder") volunteer = Volunteer.new(email: "email@example.com", password: "changeme", password_confirmation: "changeme", assigned: true) volunteer.admin = true volunteer.regions << region volunteer.save!
$ bundle exec rake db:sample_region
Note: Running the
db:sample_region rake task will create a new
Region in your database and populate it with a bunch of random data (volunteers, donors, recipients, schedule chains, etc.). For more info on what exactly is created, see
Note: Region admins will be created for the new region with email addresses based on the region's name. For example, if the region name is
San Francisco, the created region admins will have email addresses:
firstname.lastname@example.org, etc. Their passwords will all be
db:sample_region rake task does not create any
Log records, so you'll have to run the rake task to generate logs based on the newly created schedule chains:
bundle exec rake foodrobot:generate_logs.
The current production version is hosted via Heroku, collaborators can push to/pull from Heroku once their repository has been setup. To do this, first install the Heroku tool belt, then add the remote git location:
$ heroku git:remote -a boulder-food-rescue-robot
You can pull a copy of the current live database with these commands ('make datasync' will also do this):
$ heroku pg:backups capture $ curl -o latest.dump `heroku pg:backups public-url` $ pg_restore --verbose --clean --no-acl --no-owner -h localhost -U robot_user -d robot_db latest.dump
Having trouble with logs not being generated for schedule chains?
- Check that ALL locations (also called donors) are valid. Often times this causes logs to not be created for current shifts/schedule chains being run.