In the two weeks since I last wrote, we have completed two projects. These projects, unlike our first project, were team-based, both undertaken with the same team members.
One of the most difficult parts of an open-ended project like this is… where to start (hence the uber-relevance of “greenfield”). The possibilities are endless. My team started with a day to brainstorm any and all possibilities, working down to a fully-fleshed out, reasonably scoped final project idea.
The problem we chose to tackle was the confusion of group trip planning. When planning, various things are discussed (Where to stay? When to go? Which airline are you taking? What restaurants should we go to?) through various different channels (Facebook group message? Text? Email? Phone call?), involving various different travelers. We chose to prototype a solution that allows travel companions to use a single trip- and task-oriented chat forum to consolidate topical information and conversations.
Our stack consisted of PostgreSQL, Express.js, React.js, and Node.js. For more detail, check out our project documentation & wiki.
After closing out the Greenfield project, each group chose another group’s project/codebase for a “legacy” project. The task was to dive into an unfamiliar codebase and find ways to expand and add user value. We chose to inherit a fun project called CinemaPlate. Their stack consisted of PostgreSQL, Angular.js, Express.js, and Node.js.
The app provides a pairing of a random movie (culled from a Reddit Top 100 list), and a delivery restaurant, based on a user’s location.
animated gif showing the functionality of the CinemaPlate app
The app works with several different APIs — the Reddit API, MovieDB API, and Yelp API. In the end, the user receives a randomized pairing of a movie (linked directly to Netflix) and a restaurant (linked directly to Eat24).
We chose to embrace the anonymous approach — no account needed, any user could visit this site, provide their location details, and immediately receive pairings — and add enhancements for that user: in-site trailer viewing and restaurant previewing, smart zip code autocomplete, the option to “lock” one selection and randomize the other, etc. Check out our fork and associated listing of the enhancements we added in our documentation here.
It’s hard to believe, but we are now in the final stretch, and tomorrow we begin work on our thesis project — a longer-term, capstone project. I’m excited to get to work with a new group and see what’s next!