Software and n-tier implementations
GeoMedia's unique data server technology allows the GIS community to select and connect to a diverse reservoir of software data. The various owners (the Doers) of the GIS data maintain their data in their chosen software. GeoMedia implementations reside in the Business Logic tier. Here GIS objects and business objects access and manipulate data in its native format.
N-tier architecture lowers your administrative cost of client PCs, distributes your GIS investment around the enterprise, and provides scalability for the growth of your system. Each tier is virtual and may represent one or more machines allowing you to distribute the workload, dividing your GIS "business rules" from your database server.
With implementations built on these virtual tiers, people can access and analyze data in ways never before possible. And organizations can create the most economical mix of people, hardware, and software.
Software/N-tier architecture
What to look for in an Internet GIS
The data
- Don't convert it; don't translate it.
An Internet GIS should be able to access and manipulate the data in its native format. Data is separated from the business logic and presentation medium, so the ultimate consumer is free to use data without having to worry about its format or logic.
- Provide access to different GIS data formats
People need to access to many different GIS data formats at one time. An Internet GIS needs to provide the ability to view data from these different formats and/or projections in one view.
- Provide live access
Live access means on-the-fly transformation from different coordinate systems, projections, and datums. It also means the ability to work with real-time data (traffic volumes, temporary route blockages, water/tidal levels, current service vehicle locations, and more).
- Combine Raster and Vector data
Raster and vector should both be integrated in the same map. And they both should be queryable. That allows the user to select and "swap out" raster backdrops in the map.
The client
Clients with nothing more than an Internet browser on their machine should have access to advanced spatial and network analytical capabilites.
- Perform real-time GIS analysis.
For example, the user can analyze vector features over a raster of soils and then replace the soil raster with a vegetation raster without regenerating the entire map.
- The smart map itself should be programmable
The user should be able to manipulate the map presentation with client-side scripting
- Accept client-side data
Data supplied by the client side (either as points read in or as a click-on in the map) should be translated into geometry. This geometry can be displayed as a feature object, used for spatial analysis, and even stored as data in a server-side database.
- Provide application intelligence
The user can navigate the map (pan, zoom, re-center, and re-project) without compensating for losing the pixel-to-point relationship that is required with static graphics to make them appear to be interactive graphics.
- i> Provide Network routing
Typical routing capabilities include point-to-point-point routing, proximity analysis, allocation area (isoline) analysis, and dynamic segmentation. Dynamic segmentation locates a specific point(s) along a sparsely attributed linear element AND associates attribution from another data source (even if this source does not contain spatial data -- this includes some spreadsheet applications).
- Display attribute data
Display extracted data from any of the data sources on-the-fly as graphics text in the map. This means that the application can create dynamic labels and text features in the map WITHOUT having to create and store graphic text features.
The programming
- Look for OLE/COM
Internet GIS systems benefit greatly from re-usable components within an OLE/COM programming environment. Thus, custom applications can easily be built using industry-standard tools, and you are able to take advantage of Microsoft tools for scalability and load balancing.
- You can integrate COM objects with other objects on the server side. For example, take the data from a filtered record set and output it to spreadsheet and charting objects. Or, conversely, input data (such as point coordinates) from a spreadsheet, text file, and so forth, and project them as geometry in the map.
Summary
Openness of data is the key to making the Internet a truly useful environment for sharing information. It benefits both the providers and the consumers of data.
As technology advances and communication rates increase, we expect to see thin clients becoming more powerful. These client seats will be supplied with data from a variety of servers. Being able to roam the Web to search for data and seamlessly incorporate it into a map created from other sources was - until now - an unachievable dream. Today it is an attainable reality.
Being able to permit anyone to interrogate a public database is today's expectation for Internet sites. Tomorrow, the expectation will be much more interaction with the database by the thin clients.
The broad use of fast, up-to-date, reliable maps on the web will drive this industry to provide a continually improving quality of information. Exceptional new tools will not only improve the quality of communication; they will dramatically improve the quality of decision-making throughout the enterprise.