ABI CellML Meeting Minutes, 20th October 2010

Present: Dougal Cowan, Andrew Miller, Tommy Yu, Randall Britten, Yubing Shi

This week's agenda:

1) OpenCell development - enabling the COR-like view (Alan)

Discussion:

  • Randall: It would take a lot of work to get the COR-text view in OpenCell to be as useful as COR, and there is not much point enabling it as-is.
  • Andrew: there appear to be some bugs in the COR-text view, such as memory leaks.
  • Randall: still to hear back from Mike about the COR-text testing.

Planning:

  • Randall: up to Alan, but probably not worth releasing OpenCell 0.9 with the current code-base to enable COR-text view.

2) CellML API - binary distribution (Alan)

Discussion:

  • Alan: would like a "binary" release (on Windows, dll plus header files) of the API. This being available would encourage more people to use the CellML API. It would be good if these binary releases were available as snapshots and official releases.
  • Andrew: there are a large number of permutations of possible binary releases. How beneficial would having these binary releases be?
  • Randall: the suggested options for configuring the API build could be used for the binary releases.
  • Andrew: turning on all options including bindings could be a good option for binary releases.
  • Andrew: it might be worth trying to get the API into Linux distributions.
  • There was a discussion about how to distribute API binaries for the three platforms.
  • Randall: look at other API distributions and see what they do.

3) CellML API - update

Progress:

  • Andrew: continuing work on TeLICeMS. The lexer is now mostly ready, and has been tidied up. Working on the parser next.
  • Randall: Lucian Smith and Jonathon Cooper might be interested in the parser.

Planning:

  • Randall: Yubing sent us a model that required a reset rule, which could be something to concentrate on while Yubing is here.
  • There was a discussion about reset rule problems.

4) CellML core specs - update

Progress:

Planning:

  • Andrew: ModML: planning to work on the Tran et al. cardiac muscle model, recommended by Edmund Crampin because it contains both biochemical and physics modelling. Andrew will probably work first on the Rice 2008 model on which the Tran et al. model is based.

5) Repository contributions - update

  • No update.

6) PMR2 development - update

Progress:

  • Tommy: refactoring code and building test cases on the workspace side. Enabling different types of file rendering under the workspace view, which would allow creation of an exposure-like view on a file based-basis. File types (eg images, other custom file types) could have their own renderers.

Discussion:

  • There was a discussion about exposure "previewing", which is currently provided by the fact that new exposures are private by default.
  • Andrew: custom JavaScript is a security risk.
  • There was a discussion about Andre's CellML Simulator exposure type, which uses custom JavaScript. A secure way to do this would be to have the JavaScript in PMR2, under Tommy's control.
  • Dougal created a FieldML exposure.
  • Randall: it will be common to have files larger than 10MB in FieldML workspaces. There was a discussion about large files and how to handle them.