GO-ESSP CF outcomes and areas of discussion

    * CF Web site
          o official CF Web site taking shape at PCMDI
          o Kyle Halliday is setting it up (1/2 time position)
          o the site will include Trac -- issues will be raised, tracked
            and closed through this system with a complete audit trail
            of discussions
          o the site will support various email lists attached to Trac
            discussion threads
    * Moderating CF technical discussions
          o _CF standard names_ discussions will also be hosted from the
            PCMDI site.  *Alison Pamment* (remotely from BADC) will
            moderate (a 1/2 time position for her)
                + moderated discussions will address the possibility of
                  switching to a multi-level naming scheme
          o _Unstructured grid issues and "gridspec"_ issues will also
            be hosted on this site -- *Tom Gross and V. Balaji* will
            jointly moderate
          o _Other CF technical discussion_ areas will for now be
            self-moderated *by the current CF leads (?)*
          o _Past issues that have been raised_ -- at a later time
            (after time to become familiar with the technical issues)
            Kyle or (someone) will (or may as time allows) mine the past
            emails to summarize them in Trac issues
          o _In-situ standards_ (collections of time series, profiles,
            trajectories, ...) -- the possibility of hosting discussions
            of OPeNDAP standards for in-situ data will be explored
            (so-called DAPPER conventions, which will borrow heavily
            from CF).
                + Unclear whether PCMDI management may consider this
                  outside of its scope.
                + Also possible that Unidata may get involved in this
                  through a reprogramming of funds previously allocated
                  to setting up a discussion site for unstructured grids
    * Issues for improving CF conformance
          o "profiles"
                + a recognized need for "levels of conformance" --
                  probably best addressed through "profiles". 
                + profiles will mandate which CF features must be
                  supported and particular attributes that are mandatory
                + Technical issue:  how to encode the profile in the
                  global attributes
                + Possible examples
                      # IPCC AR4 and AR5 profiles
                      # a coastal modelers' CF profile
                      # a profile for the subset of CF supported by the
                        Unidata Common Data Model
          o reference CF files
                + PCMDI will be developing reference datasets for the
                  AR5 profile /(more than this?)/
          o reference applications and libraries
                + the CMOR program will be further developed as a
                  conformance checker for the AR5 profile
                + the Unidata Java netCDF library will serve as a
                  reference implementation for the CDM profile
                      # a C version of this library is important. 
                        Possibly best done as an open source project
                        starting from the C libraries in CDAT
    * Funding/governance
          o Governance white paper was distributed/discussed
          o /(other outcomes?)/
    * Events
          o Unstructured grid CF-related public workshop to be hosted by
            Unidata October 16-18