ABI CellML Meeting Minutes, 19th October 2011

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

This week's agenda:

1) PMR2 0.5 planning discussion

Discussion:

  • Looked at tracker item 2888, the plan for 0.5.
  • Tommy: could release 0.5 with webservices and Python bindings, and 0.6 could be improved search.
  • Randall: citation metadata searching is needed as a start, then further controlled search of annotation fields.
  • There was a discussion about interim search improvement solutions before the repository models are properly annotated.
  • Mike: need to find out what people want to search for.
  • Andrew: could tweak the default search relevancy code.
  • Tommy: it is now possible to change the default Plone search.
  • There was a discussion about the exposure/view "preview" feature.
  • Andre: the usability of PMR2 is generally too low - repository access, Mercurial, search, exposure builder etc.
  • Randall: it might be possible to have workspaces hosted elsewhere, eg Google code or github.
  • There was a discussion about remote workspace support, pulling from other repository sources, etc.
  • There was a discussion about usability features (product/user focus) versus functionality (research outcomes).
  • There was a discussion about PMR2 in relation to the software integration project.
  • Randall: need to note down support for metadata from other file types as a possible feature for the plan list.
  • There was a discussion about large "binary" file support.
  • Randall: FieldML v0.4 XML files can reference external binary data sources by URL, so large binary data just needs a stable URL, and does not need to be hosted in PMR2 itself.
  • Mike: could there be another software (interface) layer between the user and PMR2 to abstract away Mercurial etc?
  • Dougal: mentioned the University of Auckland's eResearch initiative and the data fabric service.

2) Metadata spec 2.0 (Mike)

Discussion:

  • Mike: announced metadata draft/discussion spec 2.0. Most of the changes since the last version are tidying up and making the spec more internally consistent. The spec is more like guidelines, using existing standards to annotate CellML model files. XPATH URL fragments can be used to annotate objects without ids.
  • There was a discussion about using RDF to make assertions about sets.
  • There was a discussion about problems with XPATH.
  • There was an extensive discussion about annotating imports.

3) CellML website front page

Progress:

  • Dougal: CellML editorial board election results announced.

4) CellML API - update

Progress:

  • Andrew: worked on cleaner C++ language interfaces.

5) CellML core specs - update

  • No update.

6) PMR2 development - update

Progress:

  • Tommy: buildbot working for PMR2 buildout. Also did some more work on the OAuth plugin.

7) Repository contributions - update

  • No update.

8) CellML meeting purpose - Andre

Discussion:

  • Andre: what is the purpose of these meetings, and what should it be?
  • There was an extensive discussion about the content and agendas of the CellML meetings.
  • Mike: many of the agenda items might work better as breakaways.
  • Poul: we could either split into different sessions, or have the agendas organized so people can skip the bits they don't want to be here for.
  • Dougal: the agenda could have no standing items, and each week items could be suggested and people can attend or not depending on interest.
  • Andre: we need to get other groups involved in discussions about CellML to see what they want or need.
  • It was decided to change to a monthly meeting.