Geodata Committee Meeting 20060406
Revision as of 07:33, 23 August 2007 by Neteler
Public Geospatial Data Committee Meeting 2006-Apr-06
Chair: Jo Walsh Minutes: Schuyler Erle IRC: #osgeo channel on irc.freenode.org Date: 2006-Apr-07 (at least in the Americas and Europe) Time: Thursday, 17:00 UTC
- Specific discussion of data sets to be hosted at Telascience
- Deciding on baseline Geodata Metadata Requirements
- Any other business?
Schuyler Erle, John Graham, Jeff McKenna, Jo Walsh, Frank Warmerdam, Ned Horning, Perry Nacionales, Dan Putler
please add yourself ;-)
- The Public Geospatial Data Committee (i.e. 'we') plans to host data at http://osgeo.telascience.org/, courtesy of John Graham and the SDSC.
- Agreed we should start by hosting +/- two showcase data sets, one raster, one vector.
- Agreed we should host these datasets in their original formats first, before adding derivative/value-added versions.
- Agreed we should host download via HTTP, download via BitTorrent, and OGC Web Service access.
- Agreed we should peer datasets through GeoTorrent. Noted that they have data format requirements for hosted data sets that may not meet our needs -- e.g. VMap0. Frank Warmerdam will speak with Richard Orchard of GeoTorrent about this.
- Schuyler volunteers to sketch out a plan for hosting
twothree initial datasets
- Start with VMap0, Blue Marble NG, SRTM30
- Host WMS/WCS via MapServer, WFS via GeoServer
- Provide human- and machine-readable metadata for all three
- Concern over metadata standards like FGDC that may be irrelevant or raise too high a barrier to contribution by users.
- Recommended we look at MetaLite, a Windows program for creating basic FGDC metadata.
- http://edcnts11.cr.usgs.gov/metalite/help/fgdc.html lists "the smallest complete "core" of the FGDC standard that would still give sufficient basis for creating a clearinghouse"
- Discussion of the basic minimum of metadata attributes to make a data set useful
- File format
- Text description
- Source URL
- Action Items
- Jo agrees to work out what is already expressed in very common metadata namespaces like Dublin Core and FOAF versus what may need to be expressed in a custom or domain-specific (e.g. GML) namespace (such as representing projection, extent data etc that can be inferred from OGR).
- John will talk to GEON participants about their work in agreeing on common metadata.