Having written about the basics of the Inspire initiative, it’s time to write more about the means by which Inspire will be implemented. Before that go through this short review of the last post:
Inpire in short:
- Inspire defines several spatial data themes
- Governmental Agencies will provide their data that belong to a theme
- The data is provided with metadata
- Online services are needed for data acquisition
There are several services that are defined:
- Registry Service: Provides the data specifications and their data models of spatial datasets from Download Services for users to browse.
- Discovery Service :Enables the search of spatial datasets and services with metadata. Also enables viewing metadata.
- View Service: Enables viewing spatial datasets on monitor, and several map view functionality.
- Download Service :Enables download spatial datasets and their subsets to your computer.
- Transformation Service :Enables transformation of the spatial datasets to Inspire defined datamodels. First it will apply to coordinate transformation from national coordinate systems to European ETRS89 coordinate systems.
- Invoke Service: Enables invoking the web services.
The services work as interface services, which means that agencies are not required to create client applications, just the interfaces that the client applications use. In the following picture you can see the services, metadata, spatial data sets, service bus and applications.
- ISO: ISO 19101:2002 Reference model, …
- OGC: GML, WFS, WMS, …
Not too hard, wasn’t it? Part II was mainly for shortly presenting the web services. In the following parts of this series, you’ll see more of the services. It’s actually my favorite component of Inspire.
Inspire defines several web services: Registry, Discovery, View, Download, Transformation and Invoke services.
What is transformation? Changing a form is a good answer, and that’s what we see in everyday life. We are eager to physically transform ourselves to look better, or we want to transform the objects around us, such as an egg, milk and salt transforms to an omelet. How can objects be transformed then? It’s not just transforming them, but knowing clearly what is the original form and the resulting form. We cannot make an omelet, if we know nothing how eggs behave on a pan, and how an omelet should look, smell and most importantly, taste.
In spatial data world, things are not always going as smoothly as with cooking omelets. Occasionally omelets burn, and might taste terrible, but I say that transforming spatial data is much harder. If you are a GIS specialist and have to transform data from format 1 to format 2. You have to have a deep knowledge of the two formats and their data models. Format 1 can have its own data model totally different from the format 2 data model. They might have vastly different geometry and attribute types, that there is no way to transform them automatically. And, the poor GIS specialist has to spend most of his work time to manually fix his data, click after click.
Previously described might have been almost impossible to solve before. There are very good transformation software for spatial data including GoPublisher, FME and others. They are meant to make people’s work easier and faster. The transformation software knows the source format and the result data, and can repeat the transformation process over and over. The only problem is that how can the specialists be assured that these transformation software help them, makes their work easier, and lets them concentrate on GIS. Is it fear of the new? Having been fixing data errors click after click for 10 years might make some afraid of changes. They surely know what the two data forms, the source format and the result, are all about. They have it all inside their heads. These GIS specialists are true specialists, they surely know what they’re doing. I say, that combining all that knowledge with modern data transformation tools, would make a lot of transformation projects much shorter.