Meeting Minutes 8th June 2006

Meeting Minutes 8th June 2006
Present: Dong Zhang, Sarala Dissanayake, Poul Nielsen, Shane Blackett, David (Andre) Nickerson,  Andrew Miller

Poul suggested we change the meeting time to Monday, 10 am. The new time suited everyone at the meeting, so will take effect from next week.

Andre will be giving a talk on the model repository at Oxford soon.

The group was interested in any feedback on the electrophysiology interface, and suggested asking specific individuals at the institute about this.

Andre has received an e-mail via Catherine, from someone trying to run a model. Peter Villiger doesn't have time to look at it yet, but may get to it later. Andre queried what the process for responding to these is. The consensus was that such requests are part of model curation, and most queries relate to the curation status (e.g. whether they work). When the curation tools become available, this can be provided on the website. In the meantime, the can be responded to individually.

Shane: Has had more problems with mozCmgui versioning, due to Firefox using libstdc++.so.5, but all current environments available causing libstdc++.so.6 to be linked. This problem is also applicable to mozCellML. Shane will set up a build environment with older compilers to avoid this problem.

Testing of code, such as mozCmgui, and PCEnv, were also discussed. It was decided that they should eventually be integrated into the CMISS testing environment.

Andre: Has been trying to get Hairer and Wanner's RADAU FORTRAN code to work with code generated from Andrew's CellML code generation service. No luck yet, although he has found an integrator problem resulting from the array layout.

Andrew: Has been working on making code generation accessible from Plone. The approach he is adopting will allow users to upload models, to create temporary objects which last for 20 minutes, and access these to download code (either as HTML, or as a C code download).

Dong: Has received the Graph API experimental code from Andrew, and has looked at it. There was some discussion about where axes should be drawn (Shane thinks it should be done in the C++, while Andrew prefers for it to be done in the Javascript, and only data done in the C++). The consensus of the group was that it doesn't matter, as long as the Javascript provides a clean API to everything. Dong is to report back on the 19th June with such an API.

Sarala: Has a new (still manually laid out) SVG diagram demonstrating here new visual notation. Nodes are used to describe the relationship between some special reaction participants, such as catalysts, and reactions (which also have their own node).
Shane didn't follow the "pacman-like" notation used for catalysts, and suggested a clearer icon. He also suggested directionality of the lines would be useful. Poul thought the glyphs could give an idea of directionality, based on what part of the glyph the line entered (e.g. triangular glyphs). The group also favoured the idea that the location at which lines entered glyphs should be significant, to avoid confusing issues (like lines joining glyphs at their point of intersection, and inappropriate inferrence of meaning from the angle of lines).