3 Days and 5 Hours of Sleep

HackUmass IV was the first college hackathon I went to, and so far has also been the hackathon in which I've gotten the least amount of sleep. It all started out with a fairly long bus ride, delayed by over two hours due to traffic from the long weekend rush. We arrived on campus after dark, and made our way to the opening ceremonies which were already in progress. Nothing of too much importance happened there, as with all opening ceremonies the only thing that happened was sponsors showing off. After that was over, I and the other BU students meandered about the building trying to find a good space to set up. Which we did. To be honest, we found a really good space to set up, as it was a room with abundant space and comfy chairs. Once our laptops were plugged in and our belongings were strewn about the room, we headed over to the sponsor booths to snag free swag, as is customary in hackathons. However, when we returned to our roost, pockets heavy with loot, we were informed that the room that we had chosen was intended to be a sleeping space for participants who believed in taking care of their health as opposed to staying up all weekend, and we were forced to migrate into a new area.

It wasn't that bad, however, as the next (and final) room we found had even more space, and all of its walls contained mounted flatscreen TVs that we could connect to from HDMI cords present on every table. The center of the room even had a high powered camera that could also be viewed from the TVs, and much fun was had. Once we all settled in for the second time, my teammate Kevin Liang and I set about making our project, Qrator. The idea behind the app was to create a system that allowed for better categorization of music, a problem I encountered over the summer when I was trying to decide what songs I was going to remove from my playlist, to ensure that the overall quality was high. The problem with current methods of categorization in most music apps is that the only options available are a 1-5 star rating or, even worse, simply a thumbs up or thumbs down. What Qrator would do would be to allow users to create their own fields, such as "Lyrical Quality" or "Danceability," and then set their own number range for each field, and rate songs accordingly. This way, users would have a more personalized and meaningful categorization of their music, allowing them to have a better sense of their playlist. Long term, the app was to support playlist sharing, so that users could see how each other viewed and rated music. Even further, such data could be mined and analyzed to gain insights into what people looked for in music, and could allow music companies like Spotify, Google Play Music, or Pandora to make better recommendations. (Pandora seems the best in this case, as they have the Music Genome Project to differentiate music)

As this was my first college hackathon, I didn't make time to enjoy the side events that most hackathons offer. I don't remember all the ones that were available at HackUmass, but I remember there being a movie showing in one of the auditoriums. Alas, it had to be missed for the sake of the project. Nevertheless, breaks were necessary, as food is a necessity, and UMass Amherst has one of the best dining halls known to man. (That's right, HackUmass allowed its participants to eat at their dining halls) The food was amazing, and definitely made up for the fact that I was getting not even close to the recommended amount of sleep. Further breaks were taken as well, as close friends of mine from high school and elementary school happened to be at the university at the same time. Eventually however, I had to return to the project.

The code for Qrator wasn't really that complex, as it was a fairly standard Node.js/HTML5 webapp. We didn't have to worry too much about writing backend code, as the only thing that was needed was the Node.js server to serve the page. All the database work was handled by Firebase, so most of the project's code executed on the client side. Really, it was just a lot of tedious event hooks and CSS. Nevertheless, it took the whole weekend to finish, which says something about the finicky nature of web design (and by that I mean CSS). By the end of the project, burn-out was an all too real problem, as brief moments of productivity were interspersed with long bouts of Futurama, which we had playing on the wall TVs. At one point the other teams in the room tuned in to the Futurama goodness from their own TVs, and all was well.

When it was time to stop working and make submissions, we realized that Qrator was not nearly close to where we wanted it to be, and didn't expect much. Nonetheless, we presented it like it was going to change the world of music, and apparently that was enough, as we were informed a bit later that Qrator made it into the finalist stage. What that meant was that we would have to give our pitch once more, in front of an entire auditorium, without the benefit of a demo. That we did, but unfortunately we didn't sell it to the judges, and didn't place among the top three. (To be fair, first place went to a group who really knew how to present)

In the end however, HackUMass IV was an unforgettable experience, and I greatly look forward to HackUMass V next year. May it be as sticker-filled as the last.

If you want to see the code for Qrator, you can find it here.
If you'd like to see photos from the event, MLH has some on their Facebook page here.

The Qrator Logo