ABI CellML Meeting Minutes, 6th July, 2011
Present: Dougal Cowan, Andrew Miller, Poul Nielsen, David "Andre" Nickerson, Tommy Yu
This week's agenda:
1) Repository contributions - update
Discussion:
- Andre: are there any plans for summer projects to do with the repository?
- Poul: possibly applying for funds for a new curator. Adding metadata to models could be the basis for a summer project or two.
- Andrew: changed the SED-ML CellML model example on the SED-ML site to make it valid SED-ML level 1 version 1.
2) PMR2 development - update
Progress:
- Tommy: working on the REST API. Fixed exposure rollover. Working on a good way to perform authentication for application access. Fixed a security flaw from the Z3C form lib.
Discussion:
- Tommy: still waiting on the Plone 4.1 release before staging 0.4.
3) CellML API - update
Progress:
- Andrew: working on support for uncertain parameter models - looking at code generation, such as what an integrator would want to do to generate these numbers. Writing a mock-up in C. Looking at both numerical and realization approaches. Written a template for CGS, and started work on a template to generate inverse of the density function. SED-ML running and processing services can now run models.
- Andrew: continuing work on converting the build system to CMake. Separating language support from core.
Discussion:
- There was a discussion about which cases or methods to supply for parameter uncertainty.
- Poul: is the spec for uncertain parameters the same as SBML?
- Andrew: SBML seem to only provide realizations (sample sets).
Action item 1: Andrew - contact SBML chairperson to find out about SBML parameter uncertainty spec.
4) CellML core specs - update
- To be discussed in a breakaway.
5) Time permitting: Updates or additional items from volunteering attendees
- Andre: started a discussion about SBML vs CellML imports and relative links.
- There was a discussion about tools, their use of xpath, relative links, XML base, etc.
- Andre: discussions about features included in the CellML API need to occur.
- There was an extensive discussion about which features (eg. SED-ML support) should be in the CellML API, and which should be separated out.
- Poul: this will be discussed at next week's meeting.