curseMetrics
At Meetup, we use StatsD and Graphite to track, well, almost everything. Critical user interactions like logins and RSVPs, as well as system errors and hits to our API.
We’ve also built our own custom dashboard that tracks aggregate user stats as well as the status of bugs and product stories. In short, we’re tracking all sorts of critical information, and making it easy to view and digest.
I recently realized that we were missing a key metric. During the last internal hackathon, I created curseMetrics, which analyzes our repository for instances of profanity.
The script aggregates curses by name, by person, and by year. From this we can tell what our favorite curse is (“crap”), our most prolific curser (Will Howard), and our most curse-worthy year (2012).
It’s a simple Node.js script that scans our commit history for curses, and stores them in a temporary sqlite3 db. The report is a simple HTML page that uses flot to display data. If you’d like to create the report on your own repository, you can find curseMetrics on github.
18 Notes/ Hide
- softpress liked this
- st3v3 reblogged this from makingmeetup
- andrew liked this
- makingmeetup posted this