Jeff walks us through the genesis of Stack Overflow and how it begot Server Fault, Super User, Stack Exchange and now Area 51.
Jeff explains Area 51 and the democratic process around establishing a knowledge-based site. The guys talk about the old Stack Exchange pricing model and how Area 51’s approach is entirely different.
Kevin asks if there’s a concern that Area 51 will have a geek skew – especially since there’s likely a draw from the existing geek ghetto.
Scott K asks about bringing experts involvement into communities. For example, including Alton Brown in a cooking site.
Jon and Jeff talk about the model of selling software and the magical wonderland that is Coding Horror.
Jeff talks about the Stack Overflow API and the Stack Exchange API Contest. There are PRIZES! The guys consider when an API is necessary and what APIs can provide. Did somebody say community-built iPhone application for Stack Overflow?
Jon asks if the Stack Overflow API will support OData. Jeff answers “yes”, and then the conversation turns to talk of data analysis and the economics of Q & A sites. Jeff gets back to the OData, “the Sharepoint of sharing data on the web”, and points us to the OData web UI which queries current SO data dump.
Scott K notes that Stack Overflow is optimized for answerers and asks Jeff for his comments on the Stack Overflow Fatigue article. Jeff talks about site popularity, community/user issues which don’t occur on smaller sites, and moderation tools.
Jon asks about expertise and tag-based badges and comments that reputation is self-correcting. Jeff stresses that the site is really optimized for those who are the best communicators and not necessarily those who are most knowledgeable.
Kevin asks how Jeff responds to folks who have no chance of gaining enough critical mass to have their interests manifest into a Area 51 site. Jeff shares his thoughts on community growth and launching and supporting sites with love.
Jeff talks about open source, driving forward and evolving his sites and the problem with trying to be all things to all people.
Jeff fields Twitter questions about Stack Overflow SQL scalability, Stack Overflow’s testing and deployment story and how Jeff’s role has changed with the introduction of many new faces working on the app. This triggers conversations about hosting on the Microsoft platform, the good and bad of Bizspark, the benefits of servers/hardware being cheap and general happiness with the stack. Jeff also speaks to continuous integration and argues against unit testing all features. Jeff speaks of the pointy-haired boss, Metcalfe’s Law, the quite guy problem and how to work with distributed teams.
The show wraps with Jeff sharing his dependence on human unit tests, cheating, optimizing for the “mistake fixing” and Stack Overflow’s loose web development process.