Node Interactive North America 2016

Massive distributed computing does not have to be a complicated affair involving many administrators and complicated protocols: all you need is the browser. Setting up a distributed computing experiment needs just a bit of node.js on the server and JavaScript on the client.

However, that is when the complicated part starts. Can you predict the performance of that computing system? Since the computing nodes of such a computing systems include persons that decide how much they are going to stay in the web page and whether they are going to come back the next day, it might seem a priori an impossible hurdle to overcome.
However, usability, complex systems, and social studies can help us give a more or less precise ballpark on how to design the system to maximize the power of people as computing nodes. 

In this talk we will present how we are using the node ecosystem to create and deploy a volunteer computing system and what we have found about its performance. Being totally open in our scientific method and results is a design decision that enhances security and also, in this case, performance