Law & Order

About two years ago I was on an exchange program with another team at Groupon. I flew out to Palo Alto, checked in to my hotel, and spent two-weeks working with another team. It was an interesting experience. I learned a lot, committed some code, and made a big impact. At the end I came back to Chicago more knowledgeable about Groupon’s core systems.

There was a lot of down-time during those two weeks. I ended up watching a lot of cable TV. What I found was that for me the most comforting show on television was Law & Order. At most times day or night I could find a channel playing one of the many varieties of the show. For whatever reason it helped me relax and decompress a bit knowing that it was a near-constant. The only problem was finding the right channel.

Thus was born I started off by scraping a site which I won’t name for obvious reasons. Then loaded the data into nokogiri and developed a series of rules for parsing the data. After an evening I had something mostly working. After another day I had a domain, a basic design, and an S3 bucket with mirrors of the content.

I had planned to run a regular rake task to keep the data updated but never got around to it. But now with “Funemployment” in full swing I figured it was time to dust it off and make it properly functional. With some modifications to the encoding and feed parsing that took most of the day it’s now running on heroku and if the scheduled job runs as expected it should stay up to date.

Another nifty feature is that by scraping all of the tv schedule I can spin off dozens of other domains for many different shows. If the traffic is high enough I might just be able to cover the costs of buying more domains and hosting with some basic ads.

A bit of a silly thing to get started with for my first full week of Funemployment. But cleaning up some of my old projects seemed like a nice way to spend some time.