Unidata

• different clients have different requirements
• requirements vs optional information
  • optional means possibility of missing info
• should be an extension of THREDDS to FGDC
• consider GCMD relationship of standards quantities for THREDDS
  • key words
  • mapping controlled vocabulary via GCMD
• URL permanence - research is required; no redirection; need traversal tools to look for broken links
• SourceForge facilities on NSDL
• WIKI has a version control with its functionality as a collaborative too, which isl used by NSDL & DLESE
• access interface method: realtime vs archived; online vs delayed, etc.
• redundant servers would provide multiple access points
• more rather than less metadata
• human interaction needed for indepth metadata
• data providers need to be source of metadata
• DODS good at metadata
• data provider tools top priority (to help them get data on-line)
• DTD validation
• DC with edu extensions (discovery not used) not enough
• consider registry
• .Edu target audience very desirable
• investigate OpenGIS to see what catalog services it provides
• unified data model not first priority
• dataset identification (unique) should be primary focus
• data discovery primary focus of THREDDS metadata
• netCDF/DODS - abstract
• ADDE/HDF-EOS - more specific (grid, point, swath, etc)
• abstract can be used for more datasets but doesn't provide as much information (need conventions); more specific requires more work to fit datasets in (even extension of model) but once client gets data it knows how to do more things with it
• should the dataType values be an ontology?
• should THREDDS metadata be restricted to discovery and description or should it be a representation of a data model?
  • discovery only, the client will loose a lot of information when crossing the boundary between discovery and data access.
• multiple existing representations on discovery side (DC, FGDC, etc), multiple representations on access side (DODS, ADDE, etc).
• developing a general data model would require lots of work
• THREDDS data model and protocols-do in partnership
• THREDDS should develop a prioritized list of audiences/users and work with them
• GDS vs DODS/THREDDS aggregation
  • should only build one configuration
  • needs a THREDDS catalog
  • work being done on GDS configuration (Joe)
• Should look toward a broader audience to participate