An exercise in mind-bending vector programming, this small project showcases "full-stack" use of Q, utilizing the static web server for initial page loading, websockets for state update, and a relatively unoptimized implementation of the Conway's Game of Life algorithm worked out mostly through trial and error. Check out the explanation file which is written as "literate" Q. If this project seems interesting, you might also want to check out quagga, qcrypt and jwt.q for other examples of mostly-readable Q, the vector language from Kx Systems.
This stat optimizer was created after a debate between friends regarding which statistics (Speed, Handling, etc.) were most important in the newest release of Mario Kart for the Wii U. The first version was a simple ruby script and received much positive feedback on reddit and was actually featured in an article on wiiudaily.com! The feedback and advice I received from some of the redditors resulted in the creation of version 2, powered by D3.js and written in CoffeeScript the page became functional quickly but was still just a big table until I realized I was a single step away from displaying the correct data to the user (in this case the characters, vehicles and wheels which resulted in a specific set of stats) and after fetching some images from the excellent Mario Wiki. Recent additions include stats for DLC content and an alternate view which gives the user a much better idea of the relationship between the statistics being compared, featuring transition animations and color interpolation!Available for browsing on Github, coffeescript was used for fun.
Trailz is a social wayfinding game allowing users to follow clues between checkpoints and discover new locations at home or when visiting a new town. The app adds controls and its UI to Google’s MapView and talks to a simple Sinatra server hosted on Heroku made specifically for this project using MongoDB for persistence. Created as the capstone project for an Android class during college.
Riptide Music was a collaborative effort of Team Cuttlefish Collective to foster discussion and discovery of new music, it was designed and built over two semesters by 5 students and formed the core of the Software Engineering capstone project. Backed by Discogs for search and importing album info upon user interaction, the recommendation algorithm relied on tag matches weighted by ratings to determine which qualities of music the user enjoyed the most and would recommend other albums that shared those tags. The engine did actually recommend albums that users were familiar with but hadn't yet visited on the site, but with luck (and enough test users contributing tags) some new music was found! Source is available on github with some mockups below. Unfortunately screenshots of the final product are still missing.
The KillerRobot tab management extension eviscerates Chrome’s memory and CPU usage by terminating tab processes without actually closing them. It is open source and on Github and was inspired by Chrome’s behavior on mobile devices. Have a look at app.js in the repo to see how simple it is!