1. Sqlite Xenia + Geoserver

Currently Geoserver (http://geoserver.com) can connect directly to several popular databases, but not Sqlite that I have seen documented. Having this path would allow the usual array of formats/products available via Geoserver to also be available via a Sqlite Xenia db instance.

Currently one path might be sqlite -> jdbc -> geoserver , but haven't had a chance to experiment or document this.

sqlite -> jdbc http://www.ch-werner.de/javasqlite/

jdbc -> geoserver http://jira.codehaus.org/browse/GEOT-241

Update April 11, 2008 After reading this post saying that using sqlite in combination with jdbc is extremely slow, am considering doing an H2 (java embedded database http://www.h2database.com/ ) Xenia version which should have better jdbc perfomance, etc.

2. Sqlite in-memory performance benchmarking

Sqlite allows the amount of memory utilized to be set via the 'PRAGMA default_cache_size' command. Compare some 'typical' query response times with the db in memory (db size say around 200-500 MB) against a file(disk) based version.

3. conversion script - Seacoos netCDF to obskml

Update Dan Ramage completed this task documented at http://code.google.com/p/xenia/source/browse/trunk/obskml/NetCDFToObsKML/scout

This would provide a path for existing, future Seacoos netCDF feeds to be available via ObsKML/Xenia also.

3.a conversion script - CarolinasCoast aggregation to obskml

Charlton Galvorino did some earlier aggregations of NWS, NDBC, USGS, etc to a hash and then to a Xenia db instance in getting the CarolinasCoast project going. Instead of going to a hash, send this data to ObsKML making it available as part of a community of feeds.

If an 'IOOS' xml encoding is used at some future date by several federal backbone providers, that common xml encoding could also be used as a conversion source to obskml or just converted directly to sql for populating a xenia instance.

4. Convert PostgreSQL 'top_of_hour' trigger to Sqlite

Update This issue has been addressed

This would allow the same observation labeling functionality given by the trigger described at http://nautilus.baruch.sc.edu/twiki_dmcc/bin/view/Main/XeniaPackageV2#d_top_of_hour_integer_d_report_h to the Xenia Sqlite version

5. conversion script - obskml to obsregistry xml

The observation registry effort http://obsregistry.org utilizes a xml formatted feed/harvest. A conversion script would help supply that feed type and support that effort.

6. conversion script - obskml to georss

Update Dan Ramage completed this task documented at http://code.google.com/p/xenia/source/browse/trunk/obskml/products/georss

GeoRSS associates a description (perhaps changing with time like a weather feed) with an associated latitude, longitude and elevation. Should be straightforward to convert hourly KML placemarks with table content into separate GeoRSS location feeds.

7. documentation of enabling an OGC SOS feed utilizing pySOS + Xenia (sqlite, postgres)

Update this issue has been addressed at http://code.google.com/p/xenia/wiki/XeniaSOS

Bill Howe has developed a python script and configuration files to OGC SOS enable relational databases with 'view' queries configured appropriately http://www.oostethys.org/downloads/sos-cookbook-python

Haven't gone through the Xenia specific steps to enable this type of service but want to and provide documentation.

8. conversion script - obskml to 'IOOS' XML encoding

Update this issue has been addressed at http://code.google.com/p/xenia/wiki/XeniaSOS

An IOOS program office http://ioos.noaa.gov/program effort is underway to define an initial XML encoding for observations which can be agreed upon and utilized to share data nationally. A conversion script would help supply that feed type and support that effort.

9. Xenia 'package' or build

Update Feb 12, 2009 this issue has been address at VMwareHome

Same as the earlier idea at http://nautilus.baruch.sc.edu/twiki_dmcc/bin/view/Main/XeniaPackageImage this would be to have a 'build' version which might utilize some configuration xml files, but otherwise outputs/products/services are already setup to go with whatever data is imported into it. Something like a knoppix/debian distribution that might live on an embedded device like a gumstix http://www.gumstix.com