Geolocation-based Q&A. Ask and answer nearby questions in real-time
Click here: http://i.imgur.com/nE1ib6Z.gifv
On Point is a geolocation-based Q&A where users ask and answer nearby questions in real-time. This full-stack application used React, Redux, Node, Express, WebSockets, and PostgreSQL.
Our mission behind On Point was to provide users faster valuable responses to their questions about the environment around them. I found that users often have questions about their surroundings but struggle to find reliable answers fast enough. On Point was created to connect users with questions to users answers around them.
I decided to pursue WebSockets using Socket.io and Redux to synchronize and maintain dynamic state. We were also able to consistently track a user's dynamic location at all times leveraging PostGIS so that we can provide the right users with the right information that were in a specific location.
One major technical challenge we faced was cluttered map points at scale. I decided to pursue clustering algorithms to dynamically cluster and render map points at different zoom levels efficiently using a distance based dispersion of K.
You can play with the app live here: onpoint.me
- Keith Alpichi
- Craig Rodrigues
- Jong Kim
- Node 6.6.0
- PostgreSQL 9.6.2
- run
npm installto install dependencies - create an
.envfile within the root directory with the following key-value pairs - CLIENT_ID (for Facebook auth)
- CLIENT_SECRET (a Facebook secret)
- JWT_SECRET (your app secret)
- TEST_DB_URL (for the test DB)
- DEV_DB_URL (for the dev DB)
- run
psql -f server/db/sql/deb-dev.sqlto create and seed development DB tables
To run the application in development:
npm run dev-client // builds and watches client-side
npm run dev-server // runs and watches development serverSee CONTRIBUTING.md for contribution guidelines.