During my short time with Stormpath, I rekindled a bit of my love for doing hacked together but fun dev side projects. This was largely driven by my then-blossoming love for Sinatra, Heroku, and PostgreSQL, the lot of which made application development incredible easy and fun again.
Unfortunately, I’ve since returned to my product management pursuits, and actively developing has again become a past-time that has fallen to the wayside. But that doesn’t mean I can’t share neat concepts, along with the specifications and rough architecture, in lieu of the actual working code. Call it a Sunday “Someday”.
Ergo, meet HippoCrit.
HippoCrit is a concept I’ve had for a while now which I’ve been toying with on the side. The elevator pitch? It’s a light-weight meta Metacritic. Cue the Twitter-esque iconic hippopotamus logo.
The idea is to fully crowd-source content (e.g., movie, game, music, etc.) reviews without all the encumbrance that currently plagues Metacritic and Rotten Tomatoes. The design removes all barriers to entry for contributing in pursuit of an aggregated public consensus under the mantra of “make it easy, make it global, make it definitive”.
HippoCrit accomplishes this by accepting submissions via Twitter with no registration and no arbitration. By tweeting at the @HippoCrit Twitter handle, anyone can submit a rating that will be added to the aggregate score for that piece of content. For example, to rate The Shining, you would simply tweet “@HippoCrit The Shining -> 4/5″.
The service will parse the results, determine the correct piece of content (spelling mistakes and such will not be accepted; learn to spell) and the rating will be determined from a set pool of potential options (e.g., fractions, percentages, stars), all of which are converted into a score out of 100%.
Each handle can only vote for each piece of content once; any subsequent vote will overwrite the previous rating for that user, and a rating can only be accepted per piece of content every 24 hours.
This data will populate a database that then acts as the back-end for a Metacritic-like website which presents all the aggregation with commentary and synopsis derived from publicly-available data pulled via APIs from Rotten Tomatoes and Rovi.
The architecture of the application involves a number of components. As you can probably guess, the application will be primarily written in Ruby via Sinatra with Postgres for the DB layer and with Heroku as the production environment.
The basic components are outlined in the following architecture map:
The sections that follow dig into the components above in more detail. more →