Minutes 13 March2006

Meeting held 11am Monday 13 March 2006

Present: Shane, Carey, Karl, Poul, Travis

Apologies: Chris, Peter, Greg, Andre

An audio recording of the meeting can be "downloaded":ftp://ftp.bioeng.auckland.ac.nz/cmiss/opencmiss/meeting-13.3.2006.ogg

Shane is back, but hasn't yet had a chance to catch up.

Karl chairing again today.

Time to start doing some work. Chris and Karl have agreed on two possible subjects: solvers and matrix vector multiply. (this includes researching packages that could do this for us)

Karl thought about starting at the cellml farming out approach, but now more inclined to consider the matrix vector problem. Shane is more convinced that starting by distributing the cell models is better step in the direction we want to go.

Karl argues that by looking at solvers and data distribution, other packages, we can learn how to use mpi - which we need to implement the cellml distibution.

Discusson: These tasks could be done in parallel. Travis offered to look a little at the solver side. (Andre has also installed MPI and had a small look). It might be helpful for Karl to look at something in this area too to get more depth and discussion oing on topic.

Decision: Karl and Travis to divide up the looking at solvers and do some each, both to look at Global Arrays and PETSCI. Will not look at propriatory packages, likely to avoid GPL'ed stuff too, LGPL'ed OK. Would also be good to look for the common elements between their API's.

Poul presented the begining of his thinking on how boundaries are to be represented in FieldML, asci art version of the white board:

<pre>

eg

o-----o o---------o | | | | | | | | o-----o o o---------o | | <-T` | T-> | | | | | | | o-----o o o---------o

Ensemble

name: dimension: n child ensembles: set of ensembles of dimension n boundary: ensemble of dimension (n-1)

</pre>