Attending: Sarah, Hans, Bryan, Phil, Antoinette, Frank, Sylvia, Charlotte, Allyn
Apologes: Eric, Lois

Actions:
A. Everyone is encouraged to play with the questionnaire and let Bryan know what needs adding/fixing using the procedure in ticket 269.
A. Allyn to generate documentation for next version of CIM and let people know about it, so they can review and comment.
A. Allyn to add column in wiki page for definitions to say whether definition is in the CIM or not.
A. Bryan to discuss priorities about definitions in the October demo/release of ESG with Cecelia.

Previous actions:

A. Everyone is to start ticketing things that aren't in the questionnaire so we can make a priority list. - ongoing
A. Definitions for terms used in CV and CIM are being documented. Allyn going through list to see which he should be adding to the CIM and adding undefined terms and emailing people for definitions. - ongoing
 * A4 Eric to review and prioritise document attached to ticket 201 giving list of everything that remains to be done for the controlled vocabulary. - ongoing
 * PA3  Bryan / Charlotte to provide guidelines for testing the questionnaire - Charlotte to put instructions on wiki and email list with link. - done (ticket 269)
A. Bryan to email Sylvia and Cecelia and Curator list to get permission to use Curator/ESG icons in questionnaire. - ongoing Charlotte to take it over. - done - permission granted
A. Bryan to talk with Eric about remapping wiki roadmap and Gannt chart. - ongoing, after Eric back from holiday. Sarah to turn it into a ticket for Bryan. - done (both ticket and discussion with Eric) - will change things gradually.
A. All to review excel spreadsheet about mapping CMIP5 variables to software components as sent to list by Gerry. - ongoing - some questions for Gerry.

Decisions:
D. Data files themselves will have the CMIP5 realm in them already. We won't need to ask modellers this in questionnaire.
D. New version of CIM will be released this week.
D. Next version of questionnaire to be released Mon 10th. Current version will be unstable on 6th/7th (security work)
D. PCMDI will be using ESG software as primary gateway to CMIP5 data. ESG software will also be farmed out to other institutes as a federated data system.
D. CV lists in the CIM won't be maintained there. In future name of CV will stay in CIM but values will be served using a vocab server elsewhere.
D. Assume no comment about the definitions means acceptance - people can make comments whenever they want.

Issues:
I. Need to develop a procedure to determine where a definition is vetted. Stuff that's in the CIM should be vetted and agreed on. Who needs to approve different definitions? Procedure for filling in definitions that don't have attributes. When to change definition from submitted to discussed to approved?
I. Coupling files and conformances issue (see UML attached to Bryan's email subject: [metafor] questionnaire, couplings,conformance and boundary conditions, todays telecon sent: Mon 03/08/2009 11:54)

Next general telco - Thurs 13th August, 3pm BST/4pm CET.

- Questionnaire news and demo

Data files themselves will have the CMIP5 realm in them already. We won't need to ask modellers this in questionnaire.

Questionnaire release promised for today. Everyone is encouraged to play with the questionnaire and let Bryan know what needs adding/fixing using the procedure in ticket 269.

Metafor staff will be able to see page picking modelling groups (e.g. NCAR, NCAS) but members of the modelling groups will go direct to their group page.

PCMDI will be using ESG software as primary gateway to CMIP5 data. ESG software will also be farmed out to other institutes as a federated data system. BADC will run it too, but may augment it.

Coupling files and conformances issue (see UML attached to Bryan's email subject: [metafor] questionnaire, couplings,conformance and boundary conditions, todays telecon sent: Mon 03/08/2009 11:54)

We only manage coupling at the top level (decided at Paris).

Couplings we know are internal couplings (to another part of the model) and external couplings (coupled to an external file). External couplings are left hanging.

Yellow bits are filled out in model pane of questionnaire. Platform stuff filled out independently. Simulations in questionnaire consists mainly of conformances. Conformances and couplings are really closely related.

Simulation should list all boundary conditions and tie them to external couplings.

We have boundary condition class, but it's part of conformance. UML gives structure of the questionnaire - it will have an impact on the CIM.

User needs to define files before they start linking to them.

Next version of questionnaire to be released Mon 10th. Current version will be unstable on 6th/7th (security work)

- CIM news

CV lists in the CIM won't be maintained there - just being put in there at the moment to validate instances. Template developed in mind maps are an instance of the CIM. Definition lists define what's in an instance.

New version of CIM out this week. Updated version has definitions in it, cleaned up references to data sprinkled throughout activity package. Added ability to use geneology. Various things needed short names, long names and descriptions - lots of clean-up work.

Every class and feature in UML has a tagged value called notes attached to it. Definitions will exist in the UML file, though we won't see it in the pictures. It will also get dumped from the UML into the XMI files produced by Enterprise Architect and the schema files. The notes will also be packaged up as text files for user documentation. CIM.rtf files at whatever level in svn directory are these documents.

Allyn to generate documentation for next version of CIM and let people know about it, so they can review and comment. Eventually definitions will be in proper vocab server and governed by external body. Before we get governance set up, we need to review the definitions internally.

CIM captures current state of our thinking. Assume no comment means acceptance - people can make comments whenever they want.

- Definitions

Easier to take definitions off wiki and put into ESG publisher rather than rooting around in CIM.

Ticket 280 - list of things we need to obtain. Attributes need definition. Not concerned with values for the CV.

Document classes and attributes in CIM. In future name of CV will stay in CIM but values will be served elsewhere.

Need to develop a procedure to determine where a definition is vetted. Stuff that's in the CIM should be vetted and agreed on. Who needs to approve different definitions.

For proposing new definitions for Grids - best person to get to approve them is Balaji. (Not necessarily content, but way they're phrased)

Procedure for filling in definitions that don't have attributes. When to change definition from submitted to discussed to approved?

One approach - assume something is approved when it finds its way into the CIM documentation - can pick definitions from the CIM rtf files.

Allyn happy to go through and describe/define structural components of the CIM. Need others to describe scientific properties - some already being vetted by scientists in USA.

By October demo/release of the software all definitions need to be present, if not formalised.