Here are my notes on Chris Boesch's talk about analyzing big data in the cloud.
Edit: Here is the video: https://www.youtube.com/watch?v=ADjIt6ZbqKU
Chris is a professor teaching many IT subjects at Singapore Management University. He advertises the cool idea to let the students teach themselves at their own pace. He will be there to answer questions but he says no one ever comes back to him. Once thrown into the cold water, people just want to figure it out themselves.
He developed Singpath, a playful approach on learning Python online via a web powered tournament system. It runs on Python on Google App Engine.
He asks his students to develop games and put them on Google App Engine. Then he asks each team to play against the games of all other teams. A big spreadsheet will show which group manages to beat which other group's game which tuns out to be a great motivation for students to try harder and implement their game more thoroughly. With Tic Tac Toe, where every game should actually be a tie and the computer should never lose, this can be done nicely. When students see that other groups are beating their Tic Tac Toe, that is motivation enough to dive back into code and improve the AI.
He also uses Coderbuddy, a service that helps to deploy and test websites on Google App Engine.
Chris discovered that when teaching people how to program using the cloud, some of them instinctively start working extremely agile. This raises interesting problems because you need teach a whole new bunch of topics like unit testing or split testing. For Google App Engine there are some tools for this:
Uploading and downloading massive amounts of data into Google App Engine is simple, can be done with a few dozen lines of code.