Personal tools
You are here: Home / openCMISS / Wiki / Minutes 19 December 2005
Log in

Forgot your password?

Minutes 19 December 2005

Meeting held 11am 19 December 2005

Present: Carey, Greg, Karl, Shane, Travis, Andre, Poul

Genral: Andre questioned whether we had infact decided on the language and compilers, we have.

1 Starting work : Data structures.

  • What are our fundamental types?
  • What is the basic interface?

This wiki has some descriptions of the cmgui data structures and cm field parameters

Karl has looked in the mesh API for TSTT, interestingly it provides a way to ask a mesh if it has some feature.

Fortran 95's lack of function pointers may complicate an object oriented API .

API or Datastructures first? We define the API then implement the datastructures that resemble the public interface.

Scope (Regions)

  • Fields - Values you can calculate
  • Nodes - Values
  • Element - Interpolation of nodes, a collection of which form a field

no more grid points, data points, collocation points, images

Much discussion of what these concepts are and what they store.

  • Should location information be treated different from field information?
  • Elements could represent the FEM idea or the notion of a coordinate system embedded within another. eg an element could have topology, coordinate system and bounds, basis are then used to describe fields over it.
  • Poul has the following concepts from his research for FieldML.
    • Point - position wrt to a coordinate system
    • Coordinate system (aka element) - dimension, limits on coordinate values
    • Basis functions/Interpolation
    • Field parameters
    • For talking about mehes, the notion of a field that glues together elements is missing
    • Further info on this can be found at "FieldML on the Physiome site":
  • Shane expresses concerns, that while there are some good concepts in the FieldML work, we need something implemented in 3 months.

out of time....

2 Development framework.

  • How are we going to develop code?
  • What documentation systems, bug systems, code management systems do we want to employ?

3 Future Meetings.

  • Field operators: Should we have mathmatical operators at a low field level?