ABI CellML meeting minutes 2009-08-26

Present: Randall Britten, Andrew Miller, Justin Marsh, Tommy Yu, David (Andre) Nickerson, Adam Reeve, Poul Nielsen

Apologies: Dougal Cowan, Mike Cooling, Peter Hunter, Catherine Lloyd


Last week's action items (also includes previous week):

1.) Justin has looked at libannotateSBML

Action item 1.) Justin to contact Morgan Taschuk about libannotateSBML

2.) Update tracker item on advanced search in PMR2

This has not been done, further discussion is required as Tommy was not present at the last meeting, in which this was discussed, and we have insufficient notes from that meeting.

3.) Add VPH NoE logo to header of cellml.org site

This is not done yet, pending a decision on exactly how it should be executed. There is a cellml.org meeting tomorrow at which a decision will be made.

4.) Justin to send specs for a new Mac to Randall

Was not done, will carry over.

Action item 2.) Justin to send specs for a new Mac to Randall


This week's agenda:


1.) Discuss CellML -> Matlab paper.

A paper discussing CellML to Matlab conversion is planned for a special themed publication discussing open source software used for cardiac research.

There was a discussion about the prudence of releasing such a paper before we describe the CellML API in full. Andrew Miller is writing a paper about the API and it may be wise to ensure that this paper is published before any description of the CellML to Matlab converter work.


2.) History of CellML

Nicholas le Novere has requested some information about the early history of CellML. Poul has plans to make some inquiries to early members of the project and compile a set of notes.


3.) PMR2 update

Tommy has recently returned to the office after a month. He has done some minor updates to the skins but has nothing new to report.

Action item 3.) Tommy to organise a further meeting to discuss the structure of PMR2.


4.) cellml.org website

James has added a link to the Mailman mailing lists page under the community section of the site. There is now a regular meeting for the cellml.org web group on Thursdays at 11am.

One of the main discussion topics at present concerns how to better integrate the look and feel of models.cellml.org with cellml.org.

There has still been no progress on setting http://cellml.org to resolve to http://www.cellml.org, although it was suggested that Gareth de Walters is not the appropriate person to ask; perhaps Andrew Cantell can assist us.

Action item 4.) James to follow up on this with Andrew Cantell

5.) OpenCell update

Justin has made several minor fixes in the Linux build requested by James, such as fixing the model reload functionality.

Justin has fixed Windows snapshot generation. Empty snapshots were being generated, and it is possible that the Windows build system for OpenCell has never been functional, despite passing the unit tests.

Recent OpenCell Windows snapshots themselves are currently not functional. Andrew and Justin suspect that this may be due to a lack of backwards compatibility within the latest release of XULrunner. This is also likely to be the cause of instability within recent Linux snapshots.

There was a discussion about the Windows build system for OpenCell and the API, and the problems with the Windows build of OpenCell. Randall considers it a high priority to get OpenCell working in Windows again and requested that Andrew allocate some time to this urgently.

Alan Garny is having problems with merging in Mercurial, although the indentical commands and process appear to work in Auckland.

Action item 5.) Justin to contact Alan about merging within Hg.

Justin needs root access from ABI IT to bioeng46, the CellML development Mac.

Action item 6.) Randall will ensure that ABI IT grant Justin root access to bioeng46


6.) CellML API update

Andrew has been working on support within the API for reset rules and infinitessimal delays. He has almost finished support for top level MathML piecewise elements. He has been making these code commits to the API trunk if they pass all required tests.


7.) CellML specificaitons update

Randall noted an objection from Tommy to using the spec as it currently stands for the PMR2 metadata editor. Tommy would like to make use of Dublin Core instead of BQS. James reiterated his desire for a metadata editor in PMR2, noting that he is currently using RDF templates and the old metadata editor in PMR1, which is not ideal.