Meeting Minutes 4th December 2006

Meeting Minutes 27th November 2006
Present: Alan Garny, Jonna Terkildsen, Shane Blackett, Poul Nielsen, Andrew Miller

Andrew noted that Alan Garny had found that mdoel data wasn't being released in PCEnv, and this is apparently due to a circular reference. Andrew will debug this, and we will reconsider whether to release PCEnv 0.1 next week after another release candidate.

The group revisited Matt's suggestion to create tracker issues. Andrew noted that Peter thought we should wait until after we get more feedback from the community before we decide on it. Shane noted that for cmgui, he likes users to add issues even if they are never likely to get implemented. Andrew noted that PCNG makes this hard, as the dependency features are not very good. There is also a risk that the granularity of issues will be wrong, and PCNG doesn't track dependencies between issues very easily. Consensus was that Andrew will make a wiki page which describes these higher granularity features (i.e. a proposal for features needed for PCEnv 0.2).

Shane noted that Peter wanted Andrew to focus on PCEnv 0.2 for now, due to resource limitations (as opposed to the CellML API documentation/porting that Alan Garny and David Nickerson were keen to see).

Poul noted that Peter has produced a list of improvements to the PCEnv user interface, merging in features suggested by Penny Noble and Jonna with his own opinions on layout. Andrew will include these features in the wiki page, together with very rough time estimates.

Andrew noted progress on PCEnv / CellML API...
  • CellML API / PCEnv RC5 released, to fix the crashing bug found by Andre.
  • The parser for the MathML Input Language has now been completed.

Andrew noted that Gareth has now updated the CellML skin, and moved Matt's changes to the custom skin into the skin SVN repository. Matt had noted in the current-development page that Michael was rewriting the skin. However, the consensus was that Gareth should continue to make any skin changes that are required, and Michael can merge them in.

Shane was seen to see rough estimates of the time taken to get various compilers working, but didn't think that Andrew should spend any time trying to predict how long this would take. The consensus was that Andrew would update the roadmap with rough guesses, and also write up the current status of this work on the page about the compilers.

The group considered making non-CORBA XPCOM bindings to the CellML API. Shane was keen for Andrew to estimate how long this would take, and write it up. Andrew thinks the main issue is not the initial time to implement this, but rather, how long we would have to delay the next release to ensure any new bugs this introduced were found. Shane noted that this work could also help his mozCmgui work.

The possibility of making a JNI binding (to access the CellML API from Java, without using CORBA) was also discussed. Andrew will give an estimate for this.

The consensus was that Andrew should make two separate sections for the CellML API and PCEnv in this document, and will  the URL to the CellML discussion mailing list when it is complete.

Alan noted that he was keen for there to be a presentation about COR. Discussion on the mailing list have concluded on Wednesday 13th being a good date. Shane was keen for this to include more than just COR (Andre might want to say something, also PCEnv). Consensus was that this meeting should target people in the institute who would like to use CellML, but don't necessarily care about the internal details of CellML, and so attract a wider audience than those who regularly attend the CellML meetings. Andre is likely to be at the next meeting, so the exact agenda will be discussed then. Shane will book the room for three hours (1pm-4pm) to ensure we can fit everything in.