The group COMC is starting to develop a container that is able to take a OWS context document linking to several data files in the internet (e.g. a GML file, a coverage in GeoTIFF, etc) and compress them with the OWS context file together in a lossless compression format. This will allow to distribute copies of them and allow a local app to visualize, analyze and decompress the data locally. The approach is similar to geopackage BUT no limited to SQLite lossy transformation and supporting more data files (in principle all) like video, 3d files, GML etc.
The experiment will consist in a service (or an application) that creates the containers from a OWS Context document. The experiment will include considerations for huge containers (container interrelated files or container lists) and the ability to select which content is embedded in the container while the rest is left as references to services o public files. It will also require a client that is able to read the container and represent it on the screen.
The container format is still to be determined but a simple ZIP file or similar can be used.
An ER can provide a conceptual framework that sets an abstract container concept and the possibility to have lossless and lossy representations. In a lossy representation data transformations and reasonable simplifications are allowed. A lossless container could be a ZIP file and a lossy strategy could be the current GeoPackage approach. There are use cases were lossy can be acceptable while others where lossless is preferable and the ER will describe them
This proposal was defined in the COMC.SWG meeting Southampton.
The group COMC is starting to develop a container that is able to take a OWS context document linking to several data files in the internet (e.g. a GML file, a coverage in GeoTIFF, etc) and compress them with the OWS context file together in a lossless compression format. This will allow to distribute copies of them and allow a local app to visualize, analyze and decompress the data locally. The approach is similar to geopackage BUT no limited to SQLite lossy transformation and supporting more data files (in principle all) like video, 3d files, GML etc. The experiment will consist in a service (or an application) that creates the containers from a OWS Context document. The experiment will include considerations for huge containers (container interrelated files or container lists) and the ability to select which content is embedded in the container while the rest is left as references to services o public files. It will also require a client that is able to read the container and represent it on the screen. The container format is still to be determined but a simple ZIP file or similar can be used. An ER can provide a conceptual framework that sets an abstract container concept and the possibility to have lossless and lossy representations. In a lossy representation data transformations and reasonable simplifications are allowed. A lossless container could be a ZIP file and a lossy strategy could be the current GeoPackage approach. There are use cases were lossy can be acceptable while others where lossless is preferable and the ER will describe them
This proposal was defined in the COMC.SWG meeting Southampton.