Note that this roadmap is far from complete
|1.1||Minor bug fixes||January 2002||2.0||Abstract core number data type away from "float" so that application specific types may be more easily accommodated. Code built to "proper" library as opposed to just a set of object files as earlier versions did.||February 2002|
|2.1||Miscellaneous cleanup and consolidation of code base||May 2002|
The original class assumes a basic data type of `float'. This type is too restrictive for cases where only integer matrices are required, or alternatively, matrices that require higher precision in their number type.