Menu
▾
▴
Home
PROJECT JOURNAL
april, 25, 2012
We discussed about version numbering and agreed to use this pattern: [version], [release], [revision].
The alpha version (leading zero) will define the basic data structures for three-point estimation and will provide the means to execute calculations. Version 0.1.0 released.
Those basic data structures will implement list-items pattern. Each item might represent a task to be estimated.
Each item must have an id number, description and three floating point numbers. Those numbers can be expressed in hours, days, months or any other significative unit and will represent the complexity considered to be necessary to realize the task. Because three-point analisys is kind of simple, we suggested the use of some custom scale of complexity insted of a time unit in a way to avoid any sort obvious of result manipulation.
april, 26, 2012
First revision released. Source code is available in c++ only until now. ThreePointItem class is now defined. Standard deviation applied to three-point estimation will be 1.675, it will provide a confidence level of 95% according literature.
We discussed other data structures as well in order to antecipate function point analisys. It seems to be a bit more complicated because we want to release indicative, estimated and detailed algorithms with alpha version.
april, 28, 2012
DataFunctionItem is now implemented. It provides the first of the required data structures to comply with function point counting which is one of the planned features for this library. Each instance of DataFunctionItem represents one task to be estimated and it calculates complexity and function points based on value-tables from IFPUG Function Point analysis Counting Practices Manual version 4.2.
april, 30, 2012
TransactionalDataItem is another data structure required for function point measurement. It is defined now. Soon I will come up with its implementation. At this point it was necessary to review UML class model in order to provide another level of abstraction. There are few methods to generalize.
may, 2, 2012
All data structures are implemented. Before move forwared to the next release some aditional unit testing is required to eliminate any bugs that wasn´t detected until now.
