NOTE: The galeon mailing list is no longer active. The list archives are made available for historical reasons.
Hi all,My apologies being late with this summary of the CF-netCDF session at the OGC Technical Committee meetings, but the AGU meetings intervened in the week after the OGC TC. I've uploaded a draft of the summary to the OGC portal http://portal.opengeospatial.org/files/?artifact_id=37055 but I'm also including a copy of the text at the end of this message because I think there will be interest beyond those with access to documents on the portal. I'm also attaching a copy of a diagram that was drawn on the whiteboard at the session. Please let me know if you have corrections or additions to the summary. Merry Christmas to all. -- Ben ============================================================================== Summary of the CF-netCDF SWG 2009 December 10 TC Session and Teleconference Draft of December 21, 2009 A CF-netCDF SWG TC session with telecom was held Thursday, December 10, 2009 at 11:00 AM Eastern Time at the OGC Technical Committee Meeting in Mountain View, CA. *OGC Architecture Board Response to CF-netCDF Candidate Standard Submission . ** * The OAB response (*with subsequent interpretation in italics*): · Scope only describes NetCDF. Abstract Test SUITE references NetCDF. And suggest compliance with NetCDF. nScope does not match the title. *In general, the submitted candidate standard is not clear about its scope . Does it cover netCDF and the CF conventions or just the netCDF. *** · What is the RoadMap? *There should be an accompanying roadmap indicating how the CF-netCDF standardization process is to be completed.* · The reference to CF conventions to cf.org. This is not appropriate. *This is a cut and paste error in the document references. The document actually references cfconventions.org and it should be www.cfconventions.org .* · Requirements are not stated and there are not normative clauses related to the requirements. *The requirements section is necessary according to the new OGC standard specification format that is listed among the pending documents. Our impression was that we could use the old format, but will look into including a requirements section.* · OAB recommendation: Send questions to CF-NetCDF SWG and table motion for approval of RFC until next OAB meeting. *The SWG will respond to the questions and update the document as quickly as possible for re-consideration by the OAB.* Discussion of OAB Response: Most of the discussion centered on the main question of how the initial document can remain focused on the netCDF classic as the core standard while setting the framework for subsequent extensions, especially the CF conventions, but possible others as well. A number of related topics were aired, but eventually narrowed to the need to make sure the initial document clearly establishes the netCDF classic as the core standard. A motion was eventually passed to address this issue: The CF-netCDF SWG will focus the initial document on the netCDF abstract data model and the netCDF classic realization and will remove the CF constraint. The motion was made by Stefano Nativi, seconded by Rich Signell, and passed by unanimous consent. Seven of the eleven voting members were present at the SWG and one participated via telecon. Additional Discussion Topics: The SWG covered several topics: As we move forward with the CF-netCDF standardization effort, we should be mindful of the fact that other related standardization initiatives may follow, e.g., HDF, GRIB, OPeNDAP, etc. and that some of these encodings and protocols can be used in conjunction with the CF conventions. Thus, while the CF-netCDF SWG is concerned mainly with the CF in the context of the netCDF instantiation, other realizations may follow. In an attempt to clarify how the CF-netCDF elements fit together, a cartoon diagram was drawn on the board showing the relationships among the netCDF and CF abstract data models, the realization of netCDF in the netCDF Classic encoding, and other possible encodings that may use the CF conventions. There is a question of whether the CF conventions can be standardized by citing the existing CF documents as normative references. This brings up the issue of whether the CF conventions governing bodies are recognized as a standards organization. It was also noted that a CF conventions standard has been submitted to the NASA Standards Process Group for consideration. This document takes a middle ground between the two approaches: 1) simply citing the CF conventions documents and 2) taking a snapshot of CF and proposing that as the standard. A suggestion was made that different parts of CF may require different approaches, e.g., the standard names table is somewhat volatile and evolves fairly rapidly. Hence it may lend itself better to standardization by reference. Other aspects, however, are more stable and perhaps are better suited to standardization of a snapshot – e.g., coordinate reference systems and data models. The idea of having separate extension standards for the different scientific data/feature types came up again. In this approach, each data type would be proposed as a separate extension to the core CF-netCDF standard. The sensible place to start would be with gridded data, for which the CF conventions are well developed and have been in use for some time. Then others would follow as they are more clearly defined and widely adopted, e.g., the new station/point data type, trajectories, swaths, irregular grids, and so forth. Remaining Issues: A few issues were identified in earlier telecons, email interactions and in the SWG session. n The discussion with NASA regarding the copyright for materials in both standard specifications needs to be concluded. n There is a typo (netCDR instead of netCDF) that needs to be corrected. n The section that notes the relationship of the CF-netCDF standard to other OGC protocols and specifications could be expanded and clarified. Stefan will work with others to come up with a draft for improving this section.
Attachment:
CF-netCDF-other_Diagram.jpg
Description: JPEG image
galeon archives: