KML 3.0: Here within the year. What would you like in it?

Open source geospatial software evangelist Paul Ramsey writes a must-read post for those of us who are curious as to where KML is headed. He’s got minutes of sorts from the most recent Open Geospatial Consortium (OGC) Technical Committee meeting, and while the meeting may have been boring, the news coming out of it is not. For those of us with short attention spans, here are the CliffsNotes:

  • KML 2.1 is becoming part of the OGC standards process (we knew this already).
  • The OGC will begin work on codifying KML 3.0.
  • Google will continue to improve on KML 2.x, and the OGC will decide whether these changes make it into KML 3.0.
  • We should get KML 3.0 within the year.
  • In Q&A time, Google Earth CTO Michael Jones was asked whether Google being the only vendor involved in extending KML 2.x didn’t give Google too much power in the process. According to Ramsey, Jones replied that “he had approached Microsoft and ESRI to join in the standardization process of KML, and will continue to invite them to the table. In the end, an OGC process with Google and invitations to the others is better than no process at all. I find it hard to disagree with that.”

While we’re on the subject, any wishes for what should go into KML 3.0? I know what I want: A way of controlling the user experience from within the code — for example, by being able to choose whether popups appear in a tour, how fast such a tour should run, and whether a tour should start automatically. I’d also love to have better time-line functionality (along these lines), and the ability to subscribe to GeoRSS and WMS from within KML network links.

Perhaps a bit further off, I’d love a way to be able to turn on or off objects in the middle of a tour process; perhaps this could be done by repurposing the timeline functionality, so that you can set user perspectives at different points in a timeline and also have objects/layers becoming visible/invisible at timeline points. This sort of timeline would be measured in minutes of duration, of course, not absolute time, as the current timeline is in Google Earth. (This suggestion is copied wholesale from Matt Nolan’s EarthSLOT, whose tour functionality truly shines — see this demo).

Oh, and before I forget — how cool would it be if you could use KML to upload your homegrown DEMs to go with your high resolution overlays?

If you think any of these ideas are particularly bad, do let me know as well.

(See a Google Earth feature wishlist from almost exactly a year ago, and check up on what’s been implemented in the meantime. Hint: The Matterhorn is looking rather fantastic in Google Earth these days.)

8 thoughts on “KML 3.0: Here within the year. What would you like in it?”

  1. Great post Stefan. I would like to see KML support more media formats for displaying content within GE instead of having to go out to the browser to do it. If there are ways to get Flash, QuickTime, Windows media, etc. to play within the GE client interface, that would be great. Not sure if this would be possible in an XML format like KML but if so, that would be fantastic.

    Related to this, it would be great to see GE have some sort of plug-in architecture that expands functionality.


  2. What I’d like to see in a new generation of KML is support for real cascading of line and icon styles ala CSS.

    For one trail-mapping application I’m working on currently, we want to use line styles to differentiate between three variables. We use line-width styles to differentiate between trail size (single-track, double-track, gravel road, paved road, etc), line color styles to indicate the relationship of the trail to others in the area (i.e. main trail or side trail), and opacity styles to indicates whether the trail is marked with blazes or not.

    With the current state of KML instead of being able to specify one style of each variable group for a given line, we must make up a different style for every combination needed, greatly complicating the maintenance of the styles and the usage of them.

    Also, more line-style options such as dotted and dashed would be helpful.

  3. My wish list would be close to points already made;

    1- being able to pull flash/QT into GE directly,

    2- being able to control what happens in a tour. I’d like to be able to make the view change THEN turn layers off and on. In a recent project I had to resort to recording the tour as a video with loss of interactive control and quality as a result.


  4. There are lots of possibilities for the next version of KML.

    Symbolisation, custom icons and classifying of sub classes of features are just a few.

    KML is describing the content, so links like network links could be links to other media files, but the application reading the KML would have to know what to do with that file.

    This could be used to add voice, navigation and description to a route.

    Though if you parsed the KML file through google maps then your would lose that function.

    So agreeing along with Adam’s comments for more styles and customisation of lines,points and polygons.


  5. Bathymetry.

    OK, so it is a function of the database/application – not of KML – but it would still be nice.

  6. What about support for storing more than one data attribute per feature? We should have attribute tables, or some kind of XML version of that.

    Perhaps the limitation is in the geo-browser not the KML format.

Comments are closed.