ABI CellML Meeting Minutes, 27th October 2010

Present: Andrew Miller, Yubing Shi, Dougal Cowan, Randall Britten, Catherine Lloyd, David "Andre" Nickerson

This week's agenda:

1) Workshop (Andre)

Discussion:

  • Catherine: asked Peter Hunter and Poul Nielsen if it is still worth running a workshop, given the frequency of HARMONY & COMBINE meetings. There are a number of problems with the timing of the CellML workshop.
  • Andre: asked if Peter needs to be present at the workshop. Peter could attend via EVO. The workshop could be more for users.
  • There was a discussion about when to hold the workshhop, possibly in February. There was a discussion about venues, costs, and funding.

2) CellML language essay (Catherine)

Discussion:

  • Catherine: previous essays were submitted in March (CellML repository and curation essays). A language essay has been requested by Chris Taylor.

Planning:

  • Andre: maybe Andrew could write this language essay.
  • There was a discussion about authorship of this essay.

3) PMR2 development - update

Discussion:

  • Randall: started a discussion about tracker item 2588, which relates to the issue of validating users who push/commit to the CellML repository. It can be hard to tell who is making the commits if Mercurial has not been set up properly.
  • Catherine: the most relevant time to check who made commits is when an exposure is created.
  • Randall: it would make sense to check the commits/pushes made since the last exposure, and check out any strange or unknown names at this time.
  • Andre: The changes are what matter, and these have to be curated, not who made the commit/push.
  • There was a discussion about comparisons with github, Google code, etc.
  • Andre: Biomodels want lots of features in the next version of their database which already exist in PMR2. A group in Heidelberg are working on the next version of Biomodels. They can't get Python programmers, and don't see the need for a distributed VCS.
  • Randall: there will probably be problems in future with large binary data files and Mercurial.

Planning:

  • Randall: at the time of creating a new exposure, the list of names in the history should be checked for any unknowns. These should be somehow noted with who they are, and an email sent to remind people to set up their Mercurial correctly.

Progress:

  • Tommy: added layout for customised file-rendering for workspace views. This makes it possible to create custom, dynamic file views per file. Multiple view types should be possible for a single file type.
  • Tommy: security issues need to be sorted out for JavaScript exposure types.

4) Repository contributions - update

Progress:

  • Catherine: Penny and Alan have been curating heart models, fixing units and making models steady-state. Metadata is included in these models, but cmeta:ids are stripped out, breaking session files.

5) CellML API - update

Progress:

  • Andrew: further work on TeLICeMS, fixing errors. Code is being ported to work without XPCOM. CCGS changes have been made in an attempt to get Yubing's model working, which is currently getting NaNs.
  • There was a discussion about how to get Yubing's model working in OpenCell.

6) CellML core specs - update

Progress:

  • Andrew: ModML - writing up the Rice 2008 model with ModML. All the reactions are coded, and now working on the physical part. There is no repetition present in the model.
  • Andrew: there is a need to access reaction values from the physical model, which is now possible.
  • There was a discussion about the application of ModML concepts to FieldML.
  • Randall: is it now possible to numerically solve ModML models? Andrew confirmed that it is.