Hive | Devpost

Inspiration

Each of our team members has many interests that range from music to gaming - we wanted to incorporate each of these in a new kind of streaming platform.

What it does

Facilitates the ability for collaborative interaction between content creators in a healthy environment while being able to serve realtime data to all interested parties. As for design, we wanted to incorporate a "beehive" theme representing our network of users, with each tile representing a new stream.

How we built it

We utilized containerization technology to quickly prototype different builds while guaranteeing functionality across deployments> Our service consists of three primary parts, a nextjs frontend which uses a hybrid of client and server side rendering for blazingly fast initial load times and seamless transitions around the site, the gin backend which utilizes gos concurrency model to allow low latency requests while guaranteeing authorization of accessed resources, and finally an rmtp server which allows users to upload their content to be delivered across our platform.

Challenges we ran into

Creating a smooth scrollable "hive" map of hexagonal tiles to display the live information about each stream. Integration between services due to incorrect CORS policies. The docker image for rtmp required a rewrite of its default entry point due to the fact that it only supported the default config. Conflicts between endpoints due to semantics in naming conventions between languages.

Accomplishments that we're proud of

We were able to overcome the challenges above, and realize our initial stylistic and functional vision for the project!

What we learned

develop and integrate a rtmp service into our architecture creating interactive and enticing ui elements and hydrating them using multiple endpoints across origins. how to handle authorization of resources to prevent malicious users from modifying or accessing sensitive information.

What's next for Hive

Utilizing websocket APIs to allow realtime chat capabilities for large numbers of users. A way to filter or "select a hive" corresponding to the categories/tags creators input.