Created attachment 76461 [details] Rendering of 3D features using OSM2World and Blender3D OpenStreetMap has currently a growing amount of building data, that have heights and further details to do 3D processing, which is already done by some minor projects: http://wiki.openstreetmap.org/wiki/3D_Development Sadly for consumers, most Viewers are work in progress and need powerfull hardware, or are "just" static isometric map renderings, that don't cover the whole planet. But the consumers might want to have a full interactive and global 3D representation, with buildings (and maybe city furnitures and nature), as what Google Earth already offers (but of course a free tool and with free data). To enable Marble users this kind of experience, I suggest two approaches: 1. Enabling the OSM vector parser to make use of 3D map features 2. Embedding the W3DS offered by www.osm-3d.org to pull online vector tiles From my POV, I suggest the 2. for several reasons: - parsing 3D features is not easy and needs powerfull hardware - using a service would have to good splitup in frontend/backend But anyway, there needs be found a solution, on how the service offer enough performance for the whole marble comunity and of course how the 3D vector tiles can be displayed in Marble
For any questions on the 3D topic in OSM, I would be happy to answer questions or to bring you in contact with the community http://wiki.openstreetmap.org/wiki/User:!i!
I would love to see 3D models in Marble and tend to agree with digi_c's conclusion that pulling 3D data from a service is probably the better solution. However, as the maintainer of OSM2World (http://osm2world.org), I'd like to pitch my project as an alternative to osm-3d.org. Unlike the backend of the osm-3d.org service, OSM2World is open source software. It also supports a larger proportion of the attributes present in OpenStreetMap data. While unfortunately no existing 3D vector service delivers models calculated by an OSM2World backend yet - so there is no short-term "plug and play" solution that Marble could integrate -, such a service could be developed in cooperation with the Marble community according to Marble's needs. If someone is willing to take up the Marble side of this, I would be happy to work on necessary improvements to OSM2World myself.
i agree that O2W has currently a more powerfull convertion process Tordanik. But as you describe, it lacks (currently) all server backend code (as it's not it' #1 goal) and infrastructure. From a "how to get things easily done" POV, I would still prever using existing technology, but as W3DS is platform neutral, we might develop an backend that could befilled with O2W parallel?