Volume 15, Number 4, December 1999
By Donald Sawyer and Joseph King
This activity was last reported in the June 1999 NSSDC News. Since then, new material has been added to the Formats Evolution Process (FEP) WWW site, and the Formats Evolution Process Committee (FEPC) has gained some additional insights.
New material includes the following: thoughts about levels of complexity and specialization in data types and structures for storing scientific data in files , by Michael Folk (Hypernews Forum); a description of ACE SIS data in HDF 4 , by Andrew Davis (Hypernews Forum); NSSDC's use of the Common Data Format (CDF) in the archive , by Joseph King (Archive Forum).
As a reminder, the primary objective of the FEPC is to promote a community dialogue that influences the evolution of data formats toward supporting greater cross-discipline access to data and toward a reduction in the number of formats that need support within the Sun-Earth Connection community.
The FEPC has found that there is a lack of sufficient common understanding of terms and concepts to compare and contrast adequately various data formats. Therefore, they have begun to fill this void through the "complex data types..." paper from Michael Folk and comments made to it. Additional revisions are expected, and the community at large is encouraged to provide comments by using the Hypernews Forum or by submitting additional papers.
At the same time, the FEPC recognizes that data formats are only a means toward an end: the objective of understanding, accessing, and using science data effectively. The types of science data objects, such as images, time series, and spectra, that need to be carried by the data formats and the services desired by the users are significant drivers for the evolution of the formats. An initial case study from this perspective is provided by the description of the Advanced Composition Explorer (ACE) Science Information Systems (SIS) data in HDF 4 as provided by Andrew Davis. Considerable effort has been expended by the ACE project in mapping their science data objects into Hierarchical Data Format (HDF) 4. Can the community come to agree on how common science data objects should be mapped to such formats? Can they agree on a set of APIs for such science data objects? If so, this would surely reduce project implementation costs and improve science data interchange effectiveness.
As the FEPC provides additional materials, the community at large is encouraged to offer to offer their perspectives on all aspects of these issues.
NASA home page GSFC home page GSFC organizational page