ABI CellML meeting minutes 2010-03-10

Last week's action items:

• Action Item 1: Tommy to investigate and circulate a solution to prevent multiple heads in Hg. Tommy is working on Python code to check that users don't commit multiple heads. Andrew sent out an example script which can achieve this. David Nickerson's concern last week is addressed because the code gives an error message telling people how to fix the problem. This needs further testing in other Mercurial clients, like TortoiseHg.
• Action item 2: Carried over as Dougal wasn't at the meeting. This week's Action Item 1: Dougal to check the browser usage for non-auckland visitors to PMR2.

Agenda for this week:

• Agenda item 1.1 - CellML 1.2 - Poul noted that some proposals that are already described are relatively straightforward, and suggested that they be put forward as official drafts. Randall noted that Andrew was in the process of testing out reset rules.
• Agenda item 1.2 - ModML - Andrew noted that he now has completed the code to translate from ModML with units to ModML core. He still needs to build an example of a model containing physical units.
• Agenda item 4 - OpenCell - Justin noted that the two things blocking a release of OpenCell 0.7 are ensuring there are no regressions in the treeview, and Mac support, and so we are probably very close to being able to make a release. Action item 2: Justin to build a snapshot of OpenCell; Geoff to test it on Windows & Linux, Catherine to test it on Mac.
• OpenCell / CellML API installer - Andrew has been working on the installer, and it is now close to being finished, but requires more work to fix some details. It pops up dialog boxes to remind the user of how to install MSVC, as it turns out that we cannot automate the install, so the user needs to manually enter the correct install path, and needs to ensure they don't choose to reboot the system after the install. Alan suggested we use MingW to avoid the MSVC complexities; however, this is difficult due to the fact that it may not be well supported by XULRunner - and we would need to go back to producing our own builds.
• Randall noted that the paper on language export from OpenCell has now been accepted.
• There was a discussion of what the priorities should be. Alan noted that he would like to see Java bindings under Windows. Andrew noted that this is mainly just waiting on people being able to build the API at all on Windows - he doesn't expect that once that becomes easy, it will be particularly hard to build the Java bindings on Windows. David noted that he thinks that getting IDA support completed should be a higher priority.
• Randall noted that as we have now branched for the release, it should be possible to re-enable the COR-like view on the trunk of OpenCell. Action item 3: Justin to re-enable COR view
• Agenda item 6 - PMR2 - Tommy has been working on the citation & license information display on PMR2, and also towards support for models.physiomeproject.org and models.fieldml.org. This is not yet fully implemented because automated filtering to show different things for different sites still needs to be done.
• There was a discussion about the performance of the site if we move it to a VPS / cloud hosting provider in the US. The conclusion was that it would probably be a lot faster for people from Europe, and a bit slower for people in NZ. Andrew noted that in order to properly measure this, we need benchmarks on latency, packet loss, and the like for the situation at present. Action Item 4: Andrew will contact Alan with information on what benchmarks we need regarding speed from Oxford; Alan will find out the information from the servers in Oxford.
• Agenda item 2 - Metadata - David noted that Frank Bergmann had created a C# libsedml implementation, and he is working on testing it, and is in contact with Frank.
• Dagmar Waltemath expressed interest to Catherine about the model annotation process Matt Halstead came up with, which she heard about from Abishek Tiwari's talk.
• Alan is working on a basic metadata editor; currently, this is being done in C++ using Qt.
• Agenda item 3 - cellml.org: Catherine has put up PDFs and short reports, as well as photos from Tommy, on the page about the workshop. She has the recordings from Evo, but is not sure what to do with these, as they are 1 GB files, and there is no recording for the first day.
• Agenda item 7 - Model contributions - Catherine has been working on the Decker et al. 2009 model, with Geoff. There is a potential problem with calcium in the model; this is due to uncertainty about where total calcium is used, and where troponin buffered calcium is used. Catherine has contacted the original author about this. Geoff has also been working on some other models, which are not yet in the repository. Catherine has been going through 8 models from the summer students, which were committed but may not be completely finished. She also noted that Mike Cooling has put some models into the repository. She also noted that Jim Wilson expressed interest at the workshop about putting his model, which is submitted to a journal but not yet published, into the repository.