Argh, i'm running late already. So tired last night i went to sleep saying i'd get up and work 6-7. Meh.
Release planning is OK.
We're using a concept called story points which are intentionally ambiguous measures of the effort, uncertainty, and complexity of a "story." (Stories are descriptions of something required.)
So a story could be (imagine the early years of the LJ Basecode): "A user wants to change the date that is associated with a post in order to post things at the front of their reading stream." One developer might say that's a 1 because they believe the task is simply to make a UI form for a user to place the date into, another developer says it's a 5 point story because there's UI work, there's changing an API, there's removing the current database assigned date, and there's testing the sort order. In this case both developers have a clear idea about the work, they discuss, and revote the points to be 5. In this case the measurement is mostly effort.
Another story could be that "People using the foo browser want the formatting in the Rich Text Editor to behave just like in the barr browser." The developers all look at each other, Foo browser? Doesn't that only run on The Foo Device? Will they have access to a Foo device? Does anyone know how the Foo browser works? And so lots of points are assigned (say 20) not because it's four times longer than the previous story, but because there's a significant amount of uncertainty.
Maybe i need to draw vector addition diagrams up so that people can visualize a measure of uncertainty and duration.
Release planning is OK.
We're using a concept called story points which are intentionally ambiguous measures of the effort, uncertainty, and complexity of a "story." (Stories are descriptions of something required.)
So a story could be (imagine the early years of the LJ Basecode): "A user wants to change the date that is associated with a post in order to post things at the front of their reading stream." One developer might say that's a 1 because they believe the task is simply to make a UI form for a user to place the date into, another developer says it's a 5 point story because there's UI work, there's changing an API, there's removing the current database assigned date, and there's testing the sort order. In this case both developers have a clear idea about the work, they discuss, and revote the points to be 5. In this case the measurement is mostly effort.
Another story could be that "People using the foo browser want the formatting in the Rich Text Editor to behave just like in the barr browser." The developers all look at each other, Foo browser? Doesn't that only run on The Foo Device? Will they have access to a Foo device? Does anyone know how the Foo browser works? And so lots of points are assigned (say 20) not because it's four times longer than the previous story, but because there's a significant amount of uncertainty.
Maybe i need to draw vector addition diagrams up so that people can visualize a measure of uncertainty and duration.
Tags: