Meeting Minutes 5 September 2007

Present: Randall Britten, Catherine Lloyd, James Lawson, Andrew Miller, Tommy Yu (speakerphone)

Action items from last week were followed up on to identify things missed out or needing a new action item.

  • Alan Garny forwarded David (Andre) Nickerson's e-mail to the list. It turns out that the issue was actually the known issue that the CellML API imports models and exposes components instead of copying components like templates into the model. There has been some discussion on the tracker / list about this, but it seems to have slowed down now.

    • Action item 1: Randall will put this on the agenda next week when Poul is expected to be back.

  • Alan Garny also examined Jonathan Cooper's errata and noted that it was about the missing exponent in the repostory.

    • Action item 2: Andrew to open a tracker item on this so we can discuss updating the errata.

  • Catherine is still working on the best practice. The document is currently a private document in a word processor. There was some discussion about what format we should work in - Catherine is not familiar with the wiki. The options discussed were leaving it in the word processor format and fixing it when the document is finished, putting it on the Plone site under members in a wiki page until it is standardised, or putting it in Subversion in restructured text as Matt Halstead has been doing recently with the repository documents.

  • Catherine noted that the best practice document refers to example models which are in the old repository and so have an inconsistent layout.

    • Action item 3: Tommy to put example models into the Plone website (but not into the repository at this stage as it can't deal with them properly yet due to lack of citations).

New issues:

  • James has requested that a priority indication be included in the subject of Bugzilla mails. Andrew suggested using the Priority MIME header instead of putting it in the subject. Everyone was happy with this. There will be need to establish a proper mapping between MIME values in Bugzilla and those accepted by mail clients. Andrew suggested that if we are going to review this mapping, we need to think about the names of the priorities too. Andrew noted that tracker priorities are all relative, but some users have initially considered them absolute and complained when their tracker item was set to 'Very Low Priority'. Randall suggested that we maintain the status quo on this for now, and send just explain this to new users.

    • Action item 4: Andrew to establish a mapping between current priorities in Bugzilla and the priority header which gets set in Bugzilla mail headers.

  • The issue of what tracker items get sent to the mailing list was also been discussed. Andrew suggested that only new items related to specifications get sent to the cellml-discussion list. James agrees and thinks anything related to other areas like the model repository can be manually forwarded if required. The possibility of sending other mails of new issues to some other list like automated-notifications was discussed.

    • Action item 5: Randall will send a proposal we start sending new CellML specification mails on to the mailing list and solicit further opinions on whether we should do this and how.

  • James has suggested that a forum for discussing curation issues be set up. Andrew suggested that this not be discussed at the CellML meeting in case it becomes so firmly associated with CellML that it is ignored by other non-CellML groups, and thinks that such a group should be CellML and the University of Auckland agnostic to be effective. Randall concurs, and thinks that while the CellML team is unlikely to discourage such a group, it needs to be a distinct effort from CellML which CellML members can participate. Andrew suggested a newgroup or some other externally hosted group.

  • James noted he has been having problems with multi-file PCEnv sessions in the latest snapshot, and considers that this should be fixed urgently because Peter wants to demonstrate some models to another group. Andrew and Randall are concerned that problems with unreleased code are being considered urgent. It may be possible for James to get it working with the latest release. The issue was deferred to be resolved outside of the meeting. This also raised the issue of whether models which work in the snapshot of PCEnv but not in the stable version should be flagged as working in PCEnv.

    • Action item 6: James will only mark models as working in PCEnv if they work in the latest release of PCEnv, but will still try to get them working in the snapshot and make a note of such models so they can be re-checked when the release is made.

  • Tommy has fixed a problem with the simulation metadata. There has also been some discussion of the RSS feed. James is still waiting for feedback on the curation flags document. Proper testing of this also depends on site-tools. Tommy noted that the site_tools source code is in an e-mail sent to him by Michael - it seems that Michael has his own private repository for developing this which we don't have access to.

  • Andrew noted that the CCGS on the site is very old code with memory leaks that have been fixed long ago. There was discussion about what version we should update to. Updating to the CellML API version from the same time as PCEnv 0.2 would be easier as there were no interface changes until then, but at some point we also need to move on to the trunk.

    • Action item 7: Tommy to look at this and contact Andrew if he has any problems. This is something which should not go onto the live site until everything has been tested thoroughly.

  • Randall has had discussions with Alan Garny and Andrew last week about the future of PCEnv. He has also been working with Andrew to help decide what we will be doing to evaluate the candidates for the new CellML job.

  • Catherine has been using COR to validate models. She has then been fixing problems identified in the text editor. She has found the validation from COR to be a useful part of her workflow.

  • Andrew has been working on an XML editor in PCEnv. This still needs a lot more work before it can be committed. He has also been working on a new workflow to ensure Mozilla trunk changes can be merged into our Subversion version. He has written up a specification for a development system, and has spent some time preparing for the interviews for the new developer.