ABI CellML Meeting Minutes, 25th of May, 2011

Present: Dougal Cowan, Tommy Yu, David "Andre" Nickerson, Randall Britten, Andrew Miller, Peter Hunter, Poul Nielsen

Last week's action items:

Action item 1: Randall - get information about hosting of these [CellML workshop EVO] recordings.

  • Done, now hosted on BeSTGRID.

This week's agenda:

1) PMR2 issues (Andre)


  • Peter: Alan has highlighted issues to do with PMR2 URLs used for publications.
  • Andre: Human readable URLs that contain information are not necessary or achievable.
  • There was a discussion about "human readable" vs precise URLs.
  • Andre: Alan would prefer a date instead of "X time ago" in the workspace listings. This could be provided as an option.
  • Andre: a tutorial about embedded workspaces is required.
  • Andre: Alan commented that workspace pages are often missing descriptions and have an owner of "admin <>".

Action item 1: Dougal - update the models that Alan has problems with.

Action item 2: Dougal - Look at a solution to the imported models lack of descriptions.

  • There was a discussion about search features in PMR2. The possibility of removing workspaces from search results was discussed.

2) CellML API future issues (Andre)


  • Andre: questions about the security of the future of the CellML API.
  • Randall: Caton Little will be working on the API in a supporting role. If Andrew is no longer able to work on it, Caton will take over and another person may be hired to replace Andrew.
  • There was a discussion about Andrew's future work allocation.

3) PMR2 development - update


  • Tommy: integrated the subrepository listing into the file browser. There is still a navigation issue which needs sorting out. Did some back-end fixes for exposure filetype. Worked on the exposure rollover issue, which is going to be quite a big job.
  • Tommy: helped Ting with some issues around creating exposures for OpenCMISS testing examples. The use-case for these is different to anything currently in the repository.


  • Randall: Ting needs nested categories for the OpenCMISS testing examples.
  • Andre: the hierarchy should not be in PMR2, but should be provided by metadata annotations which PMR2 can present in the way the user wants it.
  • There was a discussion about how to use PMR2 to store these testing resources.
  • Randall: PMR2 does not have all of the required features for Duane's model segmentation problems. We should perhaps try to address these requirements. VPH share requires us to provide webservices, and these webservices could in turn provide some of the features needed by Duane if he were to integrate his system with PMR2.
  • There was a discussion about wider requirements for storage and sharing within the ABI, Duane's requirements, and webservices.
  • There was a discussion about getting more people in the ABI to start using DVCSs generally.


  • Randall: Regarding Ting's use of PMR2 for testing, it must be noted that the OpenCMISS codebase will be moved from Subversion to Git, and hosted on Github. This might require some OpenCMISS developers to use both Mercurial (for testing examples) and Git (for source code). However, PMR2 can be modified to support Git in addition to Mercurial. Recent refactoring work by Tommy has made this more easily achievable.
  • Tommy: Is already using Git internally for PMR2 source code.
  • Randall: Gave Tommy the go-ahead to move PMR2 source code to GitHub and completely use Git for PMR2 source code.

4) Repository contributions - update


  • Dougal: Fixed the Friel 1995 model working with David Friel's corrections. Model now has a new session file and exposure, new curation status, and the workspace includes CSV output which reproduces the paper figure. Working with David on other Friel group models in the repository.


  • There was a discussion about Google Code web-editing features, and the possibility of adding these kinds of features to PMR2.
  • It must be mentioned that early designs for PMR2 included the idea of direct editing via the web interface, and early prototypes of PMR2 had partially complete features for doing this via the "sandbox" feature.
  • There was a discussion about being able to push and push to and from other repositories from PMR2.

5) CellML API - update


  • Andrew: tested the binary release candidate 4 with a simple test program - Java and C working. Putting out RC4 of the binary SDK. working on the build system, especially MinGW.
  • Andrew: currently the cross-compile for windows on the Linux build system crashes during testing due to the libxml2 in WINE. (?)


  • Randall: happy to try RC4 on Windows. Mike Cooling might also be interested.
  • Tommy: the default install of the API doesn't provide working Python bindings.
  • Andrew: fixing the Python issue would be a new feature, and shouldn't be added to this release.

6) CellML core specs - update


  • Andrew: emailed Darren Wilkinson to find out about the SBML distribution package, and suggested some possibilities for implementing some of these features.

7) ABI software book (Peter)


  • Peter: it would be useful to have input on what technologies could be used to support creation of an electronic textbook which also should contain links to examples and tutorials etc.
  • There was an extensive discussion about the format and implementation of this book (or website), as well as the sorts of things it would need to include.
    • PDF output for printing
    • LaTeX, DocBook, Wiki, etc.
    • Format for authors to submit to editor
    • Interactive components eg. zinc viewer, dynamic graphs
    • Mathematics support


Action item 3: Dougal - look at mediawiki as a solution for the ABI software "book".

  • Andrew: there may be issues around licensing, copyright etc with a collaboratively produced book.