ABI CellML Meeting Minutes, 23rd February 2011

Present: Andrew Miller, Dougal Cowan, Poul Nielsen, Tommy Yu

Apologies: Randall Britten, David "Andre" Nickerson

This week's agenda

1) ABI summer tutorials:CellML tutorial and PMR2 tutorial

Progress:

  • Dougal: Randall has booked the level 4 lab for the tutorials - this has internet access, enabling us to use a hosted PMR2 instance instead of installing VMs.

2) Metadata draft spec - discussion

Progress:

  • Poul: check if Mike is coming to the next meeting. People need reminding that the presentation is happening. [Presentation from Mike is on riday 25th Feb at 9am in the L8 Meeting Room.]

3) PMR2 development - update

Progress:

  • Tommy: has migrated exposures to use the new separated workspace storage back-end. Has built the summer tutorial PMR2 virtual machine, which will be backup in case of the online version not being available. Working on removing the final bits of Mercurial from PMR2, authentication is being generalized. This should be complete today.

Discussion:

  • Andrew: is the VM maintained or will it be a one-off?
  • Tommy: there is a build-out script, the VM must be manually configured and populated.

4) Repository contributions - update

Progress:

  • Dougal: making new exposures of the Butera et al. II model, trying to get the session file working from the exposure. This apparently works when launched from within the ABI, it is not known where else it works. The session definitely works when loaded from a local copy.

Discussion:

  • There was a discussion about the issues with the session file.
  • Andrew: Opencell has issues referencing graphs not at the top level (imports). Alternatively the browser or ISP might be caching a bad copy of the file. If the bug is with OpenCell, there is probably nothing we can do until Alan releases a new version.

5) CellML core specs - update

Progress:

  • Poul: sent an email about identifiers - a tracker item should probably be created also.

Discussion:

  • Poul: prefer the simpler approach, consistent with the way units are handled now.
  • There was an extensive discussion about using URIs or other methods to identify types and/or units.
  • There was a discussion about ways to define types and how many types might initially be required.
  • Poul: keen to take a conservative approach, with relatively few types for 1.2.
  • There was a discussion about the differences between units and types in terms of user-definition of each.
  • Poul: it would be good to use an initial set of built in types to stimulate discussion on whether more or user-defined types are required.
  • There was a discussion about representations vs set definitions (eg. doubles vs Reals)

6) CellML API - update

Progress:

  • Andrew: working on SProS, most of the implementation of constructors is done.

Discussion:

  • Test suites and debugging are required for SProS. Also need software to use SProS to run simulations based on SED-ML.

7) CellML workshop (Andre)

Discussion:

  • Dougal: Andre would like everyone to register, or if they are not attending, let him know.