This was my first year attending the Stir Trek conference in Columbus OH. From the name and from the website, I was not quite sure what to expect. Whatever I was expecting, it certainly was not 900 people in a pristine theater multiplex. All the talks were in various theaters, with the marquees lit to match the tracks. There were several clever marketing tie-ins, too, like movie posters for the sponsors. The volunteers, the food, the sessions all were excellent. Three talks, in particular, struck me as very interesting.
David Giard (@davidgiard) gave a talk titled “Effective Data Visualization: The Ideas of Edward Tufte.” Tufte spent years analyzing and quantifying how data is visualized and displayed. Giard covered Tufte’s ideas, including concepts like the lie factor, data-ink ratio, and data density. The thing that struck me was how Giard was able to take twentieth century concepts and apply them to today’s latest problems in business intelligence. As a consultant, Giard has found many ways of spreading Tufte’s message and integrating effective data visualization within projects.
Mark Stanislav (@markstanislav) gave a talk on using the cloud for disaster recovery. A DR strategy is a perfect example of owning the base and renting the burst. Recovery environments are rarely needed, and often needed only for a short period of time. Mark’s case study featured a small business with an Internet streaming product. It’s common to see 99% uptime for smaller organizations, which means 3-4 days downtime per year. Mark’s solution had an annual operational cost of $1,184 and a per incident cost of $435. The recovery time was an hour. So for a total of $1,619 annually, the small business boosted their uptime to 99.99%. Quite impressive, and a good example of how companies can use cloud computing in lower risk areas to build competency.
It was a talk by Mike Amundsen (@mamund) that blew my mind. “It’s the future. I don’t have my jetpack. I don’t have my flying car. But at least I have a browser.” And with that, Mike walked thru developing, testing, debugging, and deploying apps directly from a web browser. The toolset was Cloud9 IDE, Node.js, CouchDB, Github, and Heroku. I remember how hard it was to get into software development as a kid, so I was simply amazed at how easy it is today. I definitely need to find a project to complete entirely on the cloud for the cloud. By the way, Mike has a book out that covers his process: Building Hypermedia APIs with HTML5 and Node.
All in all, it was a great time and I left itching to write some code, test some recovery strategies, and do some data visualizations. Thanks to all the volunteers, speakers, organizers, and sponsors.