On Apr 6, 2011, at 6:13 PM, Sean Owen wrote:

> Is anyone else's historical success rate low or is it just me? 

My experience with GSoC included successes, failures, and outcomes somewhere in 
the middle (e.g. a project was delivered, but we never integrated it in the 
codebase, and it was left to rot in the repo). Our conclusion was that the type 
of task you post for your students to do plays a very important role in this 
(the last discussion on that is here: 
http://markmail.org/message/qg2ovoiwtcp3eji5 ). Most of the applying students 
are not already active hackers on a project, so giving them tasks requiring 
massive changes to the sensitive areas of the code is one recipe for disaster, 
and there are more...

Andrus

Reply via email to