2008-03-19: RSS (and related) feeds for catalogs

From Datafedwiki

There is a need for simplifying access to catalog content. In the case of data access services, it is useful to have a direct way to query a catalog for data sets that match particular criteria (e.g., list all data sets that provide air quality point observations for the U.S.).

The DataFed catalog of data access services is defined in a wiki page using the semantic wiki extension and RDF. A RSS extension for the wiki was installed to serve the catalog entries as RSS feeds.

The DataFed catalog as an RSS feed: http://datafedwiki.wustl.edu/index.php?title=RSS_Catalog&action=feed&feed=rss

Examples of filtering the DataFed catalog feed using Yahoo Pipes:

• Point data: http://pipes.yahoo.com/pipes/pipe.run?_id=fee16b7c589e0478ee5f088b7782cf64&_render=rss&data=DataType:Point
• EPA as provider: http://pipes.yahoo.com/pipes/pipe.run?_id=fee16b7c589e0478ee5f088b7782cf64&_render=rss&data=Provider:EPA
• Emissions data: http://pipes.yahoo.com/pipes/pipe.run?_id=fee16b7c589e0478ee5f088b7782cf64&_render=rss&data=Domain:Emission

The RSS feeds can be used by external clients to access the DataFed catalog metadata.
Examples:

• NEISGEI portlet
• Web Application
• Google Gadget

Others are pursuing similar, URL-based access to lists of web services.

• Brian Wilson, NASA is proposing Service Casting Feeds:

"Besides reading wikipedia articles and their references, I've particularly found these sites useful to learn about Atom (versus RSS), AtomPub, and OPML.
A good introduction to the Atom 1.0 format is at: http://www.ibm.com/developerworks/xml/library/x-atom10.html
To learn more about the Atom Publishing Protocol, see: http://www.ibm.com/developerworks/library/x-atompp1/ http://www.ibm.com/developerworks/library/x-atompp2/
The OPML 2.0 spec is at: http://www.opml.org/
The Google Data (GData) protocol spec. is at: http://code.google.com/apis/gdata/reference.html"

• Pat Cappalaere, Vightel is interested in REST-based services for accessing metadata:

"what would you suggest doing if you had an asynchronous WPS that would send an email back to a user when the job is completed. That email would contain a url to an Atom feed. That feed would contain multiple entries based on the many resulting products. These entries would contain links to KML files. KML files would contain links to GEOTIFF products. This is all obvious to an agile user to follow but how would you suggest describing that output so it could be discovered within a catalog. :)"