&ldquo;first there was a three-legged monkey...&rdquo; tag:andthennothing.net,2005:Typo Typo 2005-12-18T03:19:08+00:00 Jonas Bengtsson jonas.b@home.se urn:uuid:b7e166f1-240b-4c44-b10c-245a5e2a9fa5 2005-10-05T21:39:00+00:00 2005-12-18T03:19:08+00:00 <p>Via the <a href="http://agiletoolkit.libsyn.com/index.php?post_id=15636">Agile Toolkit Podcast</a> I discovered the article <a href="http://www.agile2005.org/XR4.pdf">Promiscuous Pairing and Beginner’s Mind: Embrace Inexperience</a> by Arlo Belshee. It&#8217;s been a while since I&#8217;ve read such an interesting article!</p> Arlo describes how Silver Platter have experimented with different approaches to pair programming and task ownership/distribution. And most of their findings are not what you expect, and often they contradict commonly held truths. Some of them are: <ul> <li>assigning tasks to the least qualified worked better long term than assigning to the most qualified or randomly</li> <li>team ownership of the tasks (team accountability) worked better than individual ownership (individual accountability)</li> <li>task grabbing (pull) worked better than assignment (push)</li> <li>swapping pairs each 90 minutes was most productive</li> <li>for time between swaps of one hour to a day it was more productive if the one who had been with the task the longest switches task than if one stayed with a task until its completion</li> </ul> <p><a href="http://www.agile2005.org/XR4.pdf">Go read the whole article now!</a></p> <p>Oh, and that they use C++ makes it even more interesting!</p>