ABI Meeting Minutes 2008-07-09

Present:  Andrew Miller, Randall Britten, James Lawson, Catherine Lloyd, Dougal Cowan, Justin Marsh, Geoff Nunns, Mike Cooling

Apologies:  Peter Hunter, Poul Nielsen, Edmund Crampin

Last week's action items

1.)  Peter Schmiedeskamp has been away; Tommy will be talking to him about setting up a VPS today

2.)  James has added his notes on PMR2 Tracker Item 248.

3.)  Andrew has created Tracker Item 1157 on updating code export in API to allow code to be generated for models containing DAE systems.

4.)  Dan Beard is using the old Physiome Sciences Inc software CellEditor to create the CellML models of enzymes in the Biological Components Databank Catherine let him know that PCEnv and COR are now much more advanced and that he should consider using them.

5.)  Catherine sent out an email requesting abstracts for the ICBME meeting, and talked further to Poul. The funding situation is still unclear.

6.)  Andre has posted a summary of his feedback about PMR2 to Tracker Item 248.

7.)  Tommy has sent an email to cellml-discussion. He posted a link to the ABI-accessible instance of the prototype PMR2 and said that he will hopefully have an externally accessible instance available soon.

8.)  Andrew created Tracker Item 1159 to solicit documentation on differences in handling of inconsistencies unit dimensionality

9.)  Moved forward to next week.

Action item 1.) Andrew will create a tracker item to address this issue within the API's validation service so that the user is warned of the inconsistency.

10.)  Moved forward to next week.

Action item 2.)  Randall to follow up on possibility of leveraging FieldML for representing maxima/minima and finite integrals

Agenda for this week

Approval of clearing of moderation flag on team@cellml.org for Dougal

  • Clearing of moderation flag for Dougal to post to team@cellml.org unanimously approved
Action item 3.)  Andrew to remove team@cellml.org moderation flag to allow Dougal to post to the list

Catherine - Singapore meeting abstracts

  • Catherine would like all abstracts to be written by 23/07/08 for submission by 30/07/08

  • Andrew has written an abstract ('tools overview') and has solicited feedback

    • Alan has also written an abstract, with quite a different focus to Andrew's
    • Alan and Andrew will collaborate on an abstract that is more representative of both views
  • Andre will definitely be attending the meeting, so Mike is almost certain to get his ideas presented at the meeting
    • Poul therefore considers this a fairly important abstract
  • Sarala is keen to attend the meeting, and has independent funding
  • James described his ideas for an abstract, which essentially amount to a 'mashup' describing how new developments in the CellML project will be put to use to create the next-generation repository
    • Ideas include: models as 'user generated content,' curation annotations / flags, implementation of semantic annotation
    • He suggested that his presentation would be focused on applications, as opposed to the more technical aspects of particular projects that other people will discuss

Update on repository contents
  • Geoff has uploaded a CellML model for Panerai et al. 1980
  • This model contains a time-delay element, but Geoff has left it out
  • This is noted in the Model Status section of the documentation.
    • Poul and James are anxious that this alteration of the model be highly evident to browsers of the repository
  • Abhishek Tiwari has been investigating time delays in CellML and he believes that in many cases omitting the time delay from the model does not actually make a huge difference to the output.
  • He has also uploaded another lumped parameter model - Shaw et al. 2007, which uses the DAE solver in PCEnv, but at present it will not run in PCEnv due to the addition of the modulus operator in the driver function.
  • This is a problem with PCEnv, not the model.
Action item 4.)  Justin and Andrew will create a tracker item to discuss this, and work on the issue. Geoff will add a link to the model to the tracker item.
  • Geoff is also still working on debugging Dan Beard's model - he is trying to build up the model part-by-part to isolate where the errors that cause the integrator to crash are.
  • Catherine has uploaded a model by Vempati et al. 2007
  • She has also uploaded and worked on curating the Rice et al. 2008 model. Hopefully Steve Niederer from Oxford will be able to help with the curation of this model.

PCEnv progress update
  • Justin has been working mainly on UI issues and bug fixes for PCEnv

CellML API progress update
  • Andrew has been working on the API's validation service
  • Specifically, he has been checking his rules against the MathML 2.0 specification and working on rules for units consistency
  • Both Randall and Andrew have been working on the Levenberg-Marquardt solver's error tolerance.

CellML 1.2 specification progress update
  • Andrew is continuing work on his document describing the CellML 1.2 typing system

PMR2 progress update
  • Tommy fixed some security issues in the PMR2 prototype that may have been a problem when the instance is hosted externally
  • He has been talking to Andre about the possibility of supporting embedded workspaces using the Hg Forest Extension 
  • He has also been talking to Andre about how to support Andre's workflow.
  • Currently Andre uses svn, and can't directly use the PMR2 prototype without changing the way he works.
  • He has sent Tommy a .zip file of some of his work so Tommy can explore how to get it into the prototype repository. Using HgSVN is a possibility here.
  • Randall noted that he has received indication that ~5 people would like to access the PMR2 prototype instance externally from the ABI
  • Randall asked if anyone would like to test out Hg in its raw form (i.e. without Tommy's web interface,) to see how complex we might consider it to be for potential PMR2 users to work with.
    • James said he would have a go. Geoff might also have a look.
  • If James finds it easy to use, then we can reasonably concentrate less attention on the web interface, and vice versa.
  • Catherine is still concerned that biologists should be able to use a good proportion of the PMR2's functionality without having to learn how to use Mercurial
  • Mike noted that some people may not have the admin priviledges to install Hg
  • Randall noted that Windows supports more user-friendly interfaces to Hg than Linux.
Action item 5.)  James to get Tommy to show him how to use Hg.

Mac support for PCEnv

  • Justin can't begin work on getting Mac support for PCEnv up to speed yet because the development machine he is using is still running MacOS 10.2
  • Randall emailed the Institute to see if anyone had any issues with upgrading to OS10.5, and nobody did.
    • James would like to see if upgrading lets him continue importing video from the tapes of the CellML workshop, which is currently stalled due to technical issues (the computer is not picking up the camera when it is plugged in.)
  • Stephen Poon, who is aware of peoples' dependencies on OS10.2 within the Institute, is currently away
  • Poul is keen for us to upgrade to OS10.5 ASAP

Cellml.org work
  • The cellml.org group has finished their content audit, and will be meeting next Monday to do a 'flash card' exercise to identify possible high-level information architectures for the site.

Saucerman et al. 2003
  • The Saucerman & McCulloch 2003 model is still broken, but it is no longer the DAE system
  • The next thing to investigate is the fidelity of the translation of procedural > declarative code