Meeting Minutes 26 February 2007

Meeting Minutes 26 February 2007
Present: David Cumin, Jonna Terkildsen, Poul Nielsen, Tommy Yu, Sarala Dissanayake, Andrew Miller

Poul noted that Andre has taken the lead in organising the CellML workshop. He has produced a draft plan for the workshop, which he has circulated to a number of people by e-mail. He is about to publish this plan on a website about the workshop. When the site is live, it will be announced to the cellml-discussion mailing list.

Tommy has been working on a replacement metadata API for the CellML repository, as the previous one was quite buggy and had some design issues. Poul suggested that Tommy keep Matt informed of his progress, and suggested that Matt and Tommy meet tomorrow.

Andrew has made progress on the PCEnv UI milestone list produced earlier, and has completed most items on the list. Most significantly, PCEnv now supports saving of models. It also supports data export to CSV, and expanded zoom and pan features, as well as a number of performance / bug fixes. Andrew is still waiting on feedback about how we should deal with switching graphs. The group were unanimously in favour of the combo-box approach, however Peter Hunter has previously indicated he favours a tabbed approach. Poul said he will e-mail Peter to discuss this.

Andrew also noted that he starts work on his PhD on the 1st March, and will only be working 10 hours a week on PCEnv from then on. However, he will be doing some modifications to the CellML API as part of his PhD. He asked about the process by which we can accept such 'third' party contributions to the CellML API. The group was happy with Andrew working on a branch in Subversion, and then when it works, consulting with the CellML community to see if it should be included into the trunk.

Jonna suggested that combo-boxes also be used in PCEnv when setting up connections, so that component and variable mappings can be selected from a list of components / variables, rather than having to be typed in. Andrew noted that we should still support typing in variable names (as well as a drop down list) because we don't want to start putting constraints on the order in which the model must be created (for example, you might want to create connections first, and then create the components and variables).

David is working on simulations of anaesthesia for his PhD, and has created CellML models for this. He plans to create models of the entire body, including cardiovascular and neurological models. He is currently facing problems with CellML tools, because he needs to solve equations which create a cycle involving two variables. Andrew said he will look into what is going on, and how we can work around the problem.