MetterBoard 2.0: A Python Twitter Wall
A few years back I built a twitter board for my friends at Metter Media. The prototype worked well and got a lot of use, but we decided it was time to make a real product out of this concept.
Building an asynchronous streaming Twitter display board with admin controls and persistent data store was less than trivial. I chose to use Flask for my web framework as it was quick and easy to get up and running. Using Gunicorn and gevent concurrency, it took a few hours to juggling threads but I was able to set up a background process that would receive tweet data, parse the response and store it in a SQLite database. SocketIO gave me the ability to display tweets in real time, as well as removing inappropriate tweets on demand. And of course, a one line Procfile had me up and running on Heroku in no time. Overall the project took about 10 hours to complete, and I look forward to continuing to develop functionality over the next few months.
We used the new MetterBoard for the first time at Microsoft’s #CIVICTECHBOS event this past Monday. It was awesome to see the live stream of tweets above the panel member’s heads as they were discussing civic journalism in the technology age.
One of the greatest benefits to using the MetterBoard was the increase in engagement. When the audience could see their tweets integrated into the event, they were more willing to participate in a thoughtful manner. This helped get the word out for the event and also helped attendees connect with each other.
Originally published at Robby Grodin.