Distributed development isn’t a new thing. Many large software systems are developed across multiple sites. For example several teams working at different lo- cations may be working on different parts of the same software product, perhaps with an integration team somewhere with the responsibility of overseeing the assembly of each part into a cohesive product. The overwhelming majority of software produced in the open source world is created by individual hobbyist developers working from home offices or student terminal rooms. This type of development is characterised by the fact that almost every developer is physically separated and that a small number of trusted developers are responsible for integrating submissions from a wide pool of contributors. These trusted developers effectively work as editors deciding which submitted “patches” to accept, and they decide how each accepted patch will be worked into the released product. The developers rarely meet, and integration of each patch or commit happens relatively infrequently, certainly by extreme programming standards.