Now that Valentine’s Day is behind us, rather than “everything coming up roses,” it seems to me that everything is coming up 3D. At least everywhere I turn.
Hot on the heels of a very informative URISA BC seminar titled “The New Dimension in GIS – 3D Analysis” (find a good summary of the day here), I’m on my way to Montreal and Quebec City to take part in a seminar series on BIM and 3D interoperability. I’ll be presenting on a number of important BIM and 3D formats and how they relate to each other, and showing how Safe’s products FME Desktop and FME Server can play a role in greasing the wheels of data exchange between such formats. In fact, I just today heard a story of a large project here in Vancouver, BC where two firms were involved at different stages during the design process, and the hand-off between them was to redraw the plans due to incompatibility between the design systems they used. Now that is a squeaky wheel if ever I saw one.
But what really got me thinking today was Oleg Shilovitsky’s blog posting asking “can we create a 3D RSS?” Such a mechanism would provide a standards-based way to communicate modest amounts of 3D information using the very successful RSS mechanisms.
To me, it seems like there are two ways this could go. One could be to take advantage of the existing GML dialect available within GeoRSS and just embed a 3D payload inside of the GML. While this could be done today without creating anything new, it seems this would impose a Geo-aspect to the 3D RSS, which may not always be desirable or necessary in the broader context. For example, what if you’d subscribed to a 3D RSS feed of the latest automotive model announcements? – That 3D data certainly doesn’t have a spatial context to it.
The other way I could see would be for a group to just invent a new dialect of RSS, much as the Geofolks have done, and garner support for that. It may be better to choose an existing XML specification for 3D data (again, GML might suffice, or X3D, or Collada, or …). But if they decide to invent a new specification, then making a provision for a spatial reference point would be ideal so that the data could be located spatially if it made sense.
One way or another, I think that having an accepted way to syndicate 3D data is an idea which will have legs. It will be interesting to watch and see where it goes in the weeks and months ahead.