ABI CellML Meeting Minutes, 20th of April, 2011

Present: Andrew Miller, Dougal Cowan, Randall Britten, Poul Nielsen

Last week's action items:

Action item 1: Dougal - follow up with Gareth to get CL & JL removed and Randall added to Google analytics.

  • Done.

This week's agenda:

1) CellML API - update


  • Andrew: the build is getting to the point of producing binary packages for Linux and Mac, as yet it is unknown whether or not these are any good. Valgrind tests are working properly and automatically. Mac builds are all working. There are still problems with the Windows builds.
  • Andrew: binary package snapshots are available in SVN:


  • Randall: is there any way of testing these packages, or do they have examples?
  • Andrew: there is no testing infrastructure in place for these packages.
  • It was agreed that examples fall under the task of improving documentation, and that the current release will not be delayed in order to add such examples. After the release, some basic examples will be provided, and possibly an incremental release of the binaries and source provided.
  • There was a discussion about providing 64bit binaries in the future. This will be possible once ABI IT provide the required build/test VM.

2) CellML core specs - update


  • Andrew: looking at creating a skeleton for the secondary specification (working title "DAEs and events" spec)


3) PMR2 development - update

  • Tommy at HARMONY meeting in New York discussing PMR2 and related issues with overseas collaborators

4) Repository contributions - update


  • Dougal: re-exposed some models which were missing diagrams. Fixed the Li Smith exposure for Steve Niederer. Updated the liver volume mesh zinc viewer exposure to use the JSON style viewer, with materials.
  • Poul: had a message from [Marcus Owen?] about whether a particular cell differentiation model could be coded up in CellML (Collier et al. 1996), for a meeting on multicellular modelling which Poul will be attending.
  • There was a discussion about coding up this model, comparing it with existing models in the repository which use multiple CellML models via imports or scripts.
  • Andrew: this model could be done in ModML

Action item 1: Dougal - look at coding up the Collier et al. 1996 paper into CellML and comment on the practicality of doing this.


  • Poul: we need some example models using infintessimal delays (events), such as cell division models.
  • Andrew: this would be much easier with boolean variables to detect thresholds.