ABI CellML Meeting minutes 2009-06-17

Present: Dougal Cowan, Andrew Miller, Mike Cooling, Randall Britten, Julian Eden, David Long, Tommy Yu, Poul Nielsen, David "Andre" Nickerson, James Lawson, Matt Halstead
Apologies: Peter Hunter, Catherine Lloyd, Justin Marsh

Last week's action items:

1) James to get the IUPS website information on the workshop updated.
  • James has sent a couple of emails - no promotional material allowed, but the blurb can hopefully be changed
Action item 1) James re-writes the blurb, including URLs for CellML and CMGUI.

2) James to get the logs of cellml.org uptime for Peter.
  • No comprehensive logs exist to provide this information.
  • Randall suggested re-thinking the decision to move cellml.org down the road to ITS, due to the problems uncovered during discussions between Tommy and Andrew Cantell.
Action item 2) Tommy (with Andrew Cantell) to devise a plan to move cellml.org.

This week's agenda:

1) VPH Toolkit paper
  • Randall reported that Jonathan Cooper is writing a paper on the VPH Toolkit, and he wants to know if anyone from ABI wants to be a co-author. Randall agreed to organize the Auckland contribution.
  • It was agreed that Andrew should write any section about the API, so as to control any overlap with his API paper.
  • Poul suggested that Randall get some more details on the paper.

2) IUPS preparation
  • James asked if any progress had been made towards IUPS preparation; posters, talks, tutorial planning?
  • Mike said that his poster is ready, to be transported by Kenneth as Mike can no longer attend.
  • Poul suggested that James push the organization for IUPS, as we are running out of time.

3) Repository contributions - update
  • James did a few units and e-notation changes before PMR1 was frozen.
  • Andre has been committing a Mercurial workspace. During the process of finishing a book chapter on cardiac electrophyiological models, Andre has extracted models from physiome.org.nz to a single Mercurial workspace, decomposed them to CellML 1.1, and generated figures for the chapter.
  • Andre said that this work could be used as part of the tutorial.

4) PMR2 update
  • Tommy is stuck at migration; there is an issue with the keywords listing which requires manual migration.
  • Randall explained that this was due to Peter Hunter seeing the keyword categories missing from PMR2 0.1rc and wanting to see them available in the PMR2 0.1 release.
  • Peter Hunter also wants to see the model listing identical to that of PMR1, with more fields in each list entry (authors, citation, etc).
  • Andre said that continually adding features could lead to problems, and ascribed many of PMR1's problems to this process.
  • Randall suggested unfreezing PMR1 until Peter is happy with PMR2.
  • Tommy does not want to unfreeze PMR1 again, as this will lead to additional data to migrate. This will slow down the changeover.
  • Tommy said that getting things indexed (PMR1 style) with adequate performance requires polluting the codebase.

5) Website redesign - update
  • James said that new content is still being put into the templates, and that more of the index pages are being made to look something like the main home page. The site is generally progressing well.
  • Tommy said that the whole Plone software center is up and running.

6) CellML API - update
  • Andrew is still working on the buildbot setup.
  • Andrew is working on getting Windows and Mac versions working for buildbot, and is also working on documentation on how to set up build environments. Buildbot will provide automated builds on each check-in, and clean nightly builds to provide snapshots.
  • Andre asked if precompiled binary versions of the API will be available?
  • Andrew said that most developers will want to start from source.
  • There was some discussion of value of having precompiled API binaries available.
  • Randall said that at the moment we probably don't have the capacity to provide these, but that we will revisit this later.
  • Poul reported that Amir is arriving next week from UCSD to work on the Python bindings.
  • Randall reported that Sarala has spent a week getting up to speed with the Java bindings. She probably needs some idea of a basic Java app to develop using the API.

7) OpenCell development - update
  • Randall said that Justin has spent a week at Oxford, which was very productive.
  • Alan has CellML to COR-text working, and is working on the other direction now.
  • Randall highlighted the potential loss of information (metadata) in converting between CORtext and CellML. This needs to be addressed.

8) CellML Specs - update
  • Randall reported that the last meeting focused on reset rules, and reset rules with or without a delay.
Action item 3) Randall to post the picture of the whiteboard from last week's meeting to the tracker item.
  • James said that Morgan is waiting for responses from us about Saint. Morgan sent screenshots of Saint.
  • Matt discussed the problems with using the SBML annotation for CellML models - difficulty with applying annotation qualifiers.

9) Updates or additional items
  • Randall said that Jonathan Cooper had mentioned an ABI paper that could have used CellML but did not.
  • James suggested using a Friday morning seminar to help show off applications of CellML.
  • Andre said that API documentation is sorely needed to encourage CellML use.
  • Poul brought up the issue of his new version of the CellML logo; do we need more discussion, and if so, who should be included?