My favorites | Sign in
Project Home Wiki Issues Source
READ-ONLY: This project has been archived. For more information see this post.
Project Information
Members

Project Goals

This project is currently in planning mode. The goal is to develop a set of python scripts that interact with various open-source tools to generate exquisit KML files.

KML has a lot of potential. It allows you to create representations of data with a great deal of interactivity, styling options, and split data into multiple files for performance optimization. But the mashing of data, behavior, and style into one file format makes it difficult to achieve good results.

Current tools don't really take advantage of all the format's potential. Many are hampered by trying to translate existing 2d cartography done in ArcMap, Mapfiles, or SLD. This leaves out the ability to utilize networklinks, temporal features, and specify interactive behavior in layers. This type of workflow is only suitable for building SuperOverlays.

Other tools suffer from a GUI workflow that means results cannot be replicated when data changes. Or you can't copy styling techniques for use with other datasets.

Finally, many of the best existing tools are prohibitively expensive. We'd like to develop a tool that is free and open source, and has first-class support for the following features:

  • Data from a variety of formats can be read and processed (using GDAL/OGR).
  • Styling and any pre-processing is defined using a single stylesheet that can be re-run whenever data is updated.
  • Data can be split into multiple NetworkLinks easily, possibly dynamically based on file size or number of features in a Folder.
  • Supports defining style and behavior using a consistent CSS-like syntax, rather than the mix of tags located within both Placemarks and Style tags currently supported by KML.
  • Can populate Camera and LookAt tags automatically based on the data used, and these tags can be customized.
  • Can split data into different folders based on attributes.
  • Feature attributes are automatically populated as ExtendedData tags on Placemarks, and these values can be configured to map from abbreviated values in the data to full names.
  • Using the CSS syntax, it should be easy to transform and simplify data for display purposes.
  • Support for advanced regionation strategies that enable the display of very complex vector data without bogging down the Google Earth client.

Implementation Plan

Currently at Step 1.

  1. Define stylesheet format and complete full documentation for how to use it (Format documented here and here.
  2. Shop around the documentation, and gather feedback and additional use cases.
  3. Refine the documentation, and gather more feedback if needed.
  4. Develop plan for implementation, and prototype a system to transform a simple collection of point data to styled Placemarks with ExtendedData attributes.
  5. Develop full system to support more complex geometry, advanced selectors, attribute mapping, networklinks, full styling options, and advanced regionation.
Powered by Google Project Hosting