Future Development Plans Future Development Plans Funded WCSWFS

Future Development Plans

Future Development Plans • Funded – WCS/WFS support added to Hyrax – net. CDF Client Library integration • Current research – Cataloging/Discovery • Wish list – – Aggregation SQL handler File responses WMS

OGC’s WCS/WFS • NASA is currently funding development of a plugin so Hyrax can read from a WCS server – WCS data exposed using DAP • IOOS/NOAA has funded a proposal to develop a WCS and WFS interface for Hyrax – DAP data exposed using WCS/WFS

WCS/WFS: How • Using DAP 4’s DDX response • Use namespaces to categorize the attributes as belonging to DAP, WCS, WFS, et cetera • Develop an ontology to map between those namespaces (i. e. , sets of attributes) • The data models for both DAP and WCS/WFS are similar enough to enable a crosswalk using an inference engine • The ontology(ies) used will model both the Interfaces (DAP, WCS, WFS) and the response semantics.

The net. CDF Client Library • NSF funded two year effort • Move support for the net. CDF Client Library out of OPe. NDAP and into Unidata – Unidata is the author of net. CDF – Our library is currently a ‘version’ of their official library – The software will move into Unidata’s source code repository and will be an integral part of all future releases of net. CDF.

net. CDF Client Library: How • Using the Ocapi C library we developed recently • The current net. CDF CL software is written in C++ – The design will be recast in C – The libdap (C++) code will be replaced by Ocapi (C) code – CDM support will be integrated

Cataloging & Discovery • NSF funded research effort with UCSB and URI • Using a crawler to harvest metadata • Developing clustering techniques to automatically build inventories • Inventory search is closely related to data source aggregation

Inventories • An inventory is a catalog that falls between the individual data sources (DAP URLs, files, database blobs) and a list of ‘logical’ datasets • A logical dataset is what a person typically thinks of when they use the word ‘dataset’ • The inventory collects all the pieces that the computer manages and lists each one that makes up the logical dataset – My apologies for the phrase ‘logical dataset’

Aggregation • As TDS currently supports, data sources (e. g. , files) from within an ‘inventory’ are combined into a virtual data source • The virtual data source is typically ‘extended’ and choosing points along this extension is effectively searching an inventory – Example: A set of two-dimensional images are combined to form a cube with the new dimension the date/time of image capture – Choosing a range of dates/times along that new axis returns the same data as choosing individual inventory entries

Other things we would like … • SQL – A handler for SQL – Currently py. DAP and DRDS provide this support – Please contact us before using DRDS as it is no longer supported • File-type responses – We would like for Hyrax to be able to return a net. CDF, HDF 4/5 or Geo. TIFF file – The server can return ASCII CSV responses now and we have some work done on a net. CDF tool, but no funding • WMS – Since we have WCS/WFS support planned, WMS support is a natural thing to expect.