ABI CellML Meeting minutes 2009-06-24

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

Last week's action items:

1) James has sent an email about the updated blurb for the tutorial, and is awaiting a reply.

2) Tommy has had a chat with Andrew Cantell, who agrees that hosting at ITS will be a good idea. A plan still need to be written.

3) Randall has posted the picture on tracker item 1543

This week's agenda:

1) VPH NoE deliverables
  • James started a discussion about the emails from Sharon Lloyd about the VPH NoE deliverables.
  • Peter clarified that everything we are involved in is part of WP3.
  • Peter said that we should describe PMR2, OpenCell, and possibly the new website.
  • Andrew said that we need to know when the last report was so that we only cover what is new since then.
  • Peter said that the toolkit release date is the end of July.
  • Catherine said that the documents required are the annual report and the WP3 report - these both require the same information and can essentially be the same.
  • Catherine and Peter agreed to coordinate to check exactly what is required.
  • There was an extensive discussion about the Saint annotation tool and annotation problems in general.
    • Peter asked what the future is for Saint and CellML annotation; will Saint support Sarala's ontology in the future?
    • There was a discussion about Sarala's ontology, metadata editing, and CellML's specific requirements and wants.
    • Randall said that we need a tracker item for Saint/metadata editing issues.
    • Peter said that a roadmap is required so that the community can see the progress and plans.
    • Randall raised the possibility of generating a roadmap dynamically from the tracker, using some kind of tag system. This would avoid having obsolete roadmaps on the site.

2) IUPS preparation
  • A breakaway was called to discuss this.

3) PMR2 update
  • Tommy reported that PMR2 is now live. Alan has sent an email with comments about the need for better documentation.
  • Tommy does not think that the model submission functions of PMR2 are ready for public consumption.

  • Randall said that as PMR2 was basically retrieval only from the general user point of view, we need to address the ease of model retrieval.

  • There was a discussion about the use of Mercurial and public use of PMR2. Andrew suggested that we may want to encourage the use of Mercurial.
  • Randall has a list of some minor content changes to make PMR2 more obvious for model downloaders.
  • James would prefer the option to have the entire model listing on one page (to use browser text-search).
  • Tommy said that a view for a complete listing can be added.
  • There was a discussion about whether the PMR2 content should be version controlled in addition to the code. It may be that eventually this will happen - this would be desirable in the case that other people want to set up mirror or other repositories.
  • Catherine asked whether the keyword listing will be reliable - as this metadata now has to be entered by hand, people may forget to add it. This may lead to models that should be in a category not appearing in that category's listing.
  • James asked if it would be possible to search for models that are missing the keyword tag/s.
  • Tommy said that he can make a utility to find these models.
Action item 1) Tommy to make a list of models that lack "keyword" metadata.
  • James said that the current exposure creation interface is very lengthy and involved. James would prioritize interface improvements for the next version of PMR2.
  • Randall said that we need to start planning for PMR2 0.2 features, eg.
    • MathML pdfs.
    • Edit metadata via web interface.
    • Web interface for Mercurial functionality.
  • James would like a kind of "regenerate exposure" function, that uses all the old curation information to recreate a new exposure, and automatically deprecates the previous exposure.
  • Catherine raised the issue of where she has uploaded 3 variants, she can't set the curation information separately for each.
  • Tommy said that curation status is per-exposure.
  • There was an extensive discussion about curation levels in exposure with respect to variants, versions, and CellML 1.1 models that import components from different workspaces.
  • James said that we need a cohort of test models to thoroughly test the new curation flag system.
  • Tommy said that other desired features would be:
    • Additional and better exposure types.
    • Embedded workspaces.
  • Andrew asked if we need some kind of markup language for exposures - this could be required for sharing exposures between repositories.
  • Randall suggested that we have another meeting to discuss PMR2 0.2 features.
Action item 2) Tommy to arrange a meeting on Friday or Monday to discuss PMR2 plans.

4) Repository contributions - update
  • James has added the Liren-Li and Nick Smith's model, the first new model in PMR2.
  • Peter said that we need a policy to generate a session file for each new model.
  • James mentioned the difficulties with session files due to the changing interface of OpenCell.
Action item 3) James to create a session file for the Liren-Li model.
  • Catherine has completed one big cardiac model, and has nearly finished 2 others. She has also finished Kenneth's extension of the Rice et al. 2008 model.

5) Website redesign - update
  • Catherine said that she has looked at the site and thinks it looks good.
  • Randall asked whether we should freeze content editing on the old site; Tommy said that any changes can be seen in the recent changes list and these can be brought over to the new site. It may be too long before the switch to freeze the old site.
  • Poul said that we want people to look at the new content now and make comments.
Action item 4) James to send an email requesting comments and feedback about the new site.
  • There was a discussion about hosting issues and what needs to be done to get the new site hosted at ITS.
  • Peter said that arrangements for hosting at ITS should be organized immediately.

6 & 7) OpenCell development and CellML API update
  • Andrew continues to work on the build system. There is a problem with GSL on Win32; Andrew will overcome this by turning it off in Windows for the time being, leaving only CVODE for Win32.
  • All the build system work is coming along well.
  • Randall asked whether Justin should be working on the Mac build system when he returns, but it was agreed that Andrew is probably best positioned to complete the Mac work.
  • Andrew reported that Alan continues to work on the COR-text views, and that Alan has also changed the maths views to use different representations of certain expressions.
  • Randall reported that Sarala has been writing documentation on the Java API.

8) CellML specs - update
  • Randall reported that at last weeks meeting it was decided that reset rules may need at least infinitesimal delays.