[Inspire-data] Hello 2

Just van den Broecke just at justobjects.nl
Wed Nov 11 05:00:42 EST 2009


Hello,

If I understand the INSPIRE standards and the various working groups in 
ESDIN/EURADIN, the focus is on providing (web)services that comply with 
the INSPIRE standards/application schemas. Every national (mapping) 
agency (NMA) will have their data in a native format, so undoubtly 
transformations are required. This is often a very complex issue, 
especially if done in a single (transformation) step.

In order to partition the problem, and using GeoFOSS, my idea (and I 
think elaborating on Milo's mail) is to have an architecture with the 
following components:

1. PostGIS - DB schemas according/close to the INSPIRE themes/models
2. GeoServer2 (or Deegree) with Application Schema support
3. GeoNetwork for INSPIRE MD (Jeroen?)

4. ETL (to be determined) for ETL-ing data from native sources (e.g. 
Oracle) to PostGIS.
5. Visual tools for ETL ? This is at least what all closed-source 
providers advertise and we should look into.

The above approach is also called Combined Transformation in ESDIN, i.e. 
the first step is "offline" transformation using ETL (or other tools, 
like XSLT/GDAL-OGR) from native into PostGIS. The second step is 
"on-the-fly" within GeoServer2/Deegree using Application Schema support 
e.g. as part of a WFS (WMS?WPS?) service. "Application Schema support" 
is a standard facility of GeoServer2 (a.k.a. Community Schemas), 
basically a set of rules specified in XML for mapping multiple 
data/feature sources to Application Schema GML. See also
http://docs.geoserver.org/trunk/en/user/data/app-schema/index.html
I am less familiar with Deegree.

IMO components 1-3 would be a standard "Geo FOSS INSPIRE Stack", i.e. 
independent on any national native data format. I am less familiar with 
ETL suites, having used XSLT+GDAL-OGR+PGTools up to now for this step. 
There is also an OSS project in this area: http://www.esdi-humboldt.eu


best,

Just van den Broecke


Milo van der Linden wrote:
> Hi there!
> 
> I am currently inquiring technical solutions for inspire data. For this
> my primary focus is on (geospatial) ETL; from- and to- conversion for
> the INSPIRE data format.
> 
> I am investigating two projects:
> 
> geokettle:
> <http://geosoa.scg.ulaval.ca/en/index.php?module=pagemaster&PAGE_user_op=view_page&PAGE_id=17>
> 
> And spatial data integrator:
> 
> <http://www.spatialdataintegrator.com/>
> 
> Both are based upon "regular" ETL suites (Pentaho and Talend)
> 
> It is my personal opinion that using INSPIRE format(s) as a
> transportation mechanism first would be a good start. Native support for
> given formats can be something for the future *as long* as you are at
> lease capable to create or convert INSPIRE data.
> 
> What is your opinion on this?
> Do you think connecting to "mainstream" ETL is good?
> Would you prefer viewers, desktop gis or databases to handle INSPIRE
> data natively?
> Or would you be good with concentrating on "middleware" first?
> 
> I am curious about the general opinion on this topic.
> 
> Kind regards,
> 
> Milo van der Linden
> 
> 
> _______________________________________________
> Inspire-data mailing list
> Inspire-data at lists.osgeo.org
> http://lists.osgeo.org/mailman/listinfo/inspire-data
> 






More information about the Inspire-data mailing list