Today we held our second backlog selection meeting for Sprint 9. We’re running Sprint 9 as two separate sprints, one for internal conversion efforts and a second to lay the groundwork for multi-proc build* support. Since my original blog entry was getting rather long in the tooth I’m splitting it up into three parts. The next two blog entries will cover the backlog items selected for each sprint.
When looking at the two backlog lists, you’ll see some are marked as “critical” and others are marked as “target”. When we schedule our sprints we try and fill roughly 60% of our available resources with critical work items, and the remaining 40% goes to other work items that we’d like to see complete. This gives us some items that we can safely cut if the critical items go longer than we expected.
* When we say multi-proc build we really mean multiple threads, multiple CPUs, multiple cores, and multiple hamsters. It also has an impact on our longer-term plans for multi-machine (a.k.a. distributed) builds.
[ Author: Neil Enns ]