From the entry about an 3D streaming service, some pointed out, that they would be interested to work for money at their specific 3D topic. From my POV we can only boost the use of 3D tags, if we enable people to easily visualise the results and offer them a benefit for their work (rendering and sharing stills or animations of their own cities).
Maybe we can start an discussion what parts might benefit from such an code sprint and for whom it is necessary and possible?
Any ideas and comments welcome
Possible topics for a sprint (say 6-12months):
-3D streaming service (transport)
-3D DEM/height model with VGI spirit
-improving 3D clients/converters
Possible sponsors for a sprint?
-open media projecs OSMF / OSGeo / FOSSGIS / OpenData Network/ Wikimedia / OKFN / Blender foundation …
-Google Maps competitors Microsoft / yahoo / …
-GIS companies in general (ESRI, …)
-3d browsers (layar, …)
-3d content industry
-global community (crowdfunding)
There are already some promising university projects so I could imagine to get more research and manpower support for free.
Anyway this needs much better coordination. There are just too many projects and tehre is nothing really enlightning for the public (specifically talking non OSM people).
I’ve been working for a big player in geo business for a while so its really fun if you can work on this and get paid for it
At least, from there I can provide some knowledge about WebGL Map engines.
Well the problem is that research is focused on science (“inventing new fundamental things”) and not focused on the engineering process itself (“putting things together and polish up”). This results in problems, like they need to stay on this particular topic and that they can’t ship/open services to the public (due to competition and ressources) and they track topics only for a certain time.
So IF we wan’t to get an 3d stack that plays together and getting more people to contribute, we IMHO need this components:
web online client
offline native client (esp. for mobile)
3D streaming service
3D elevation service/editor
3D model service
exporters for more formats
You see, a lot of things are already in the pipeline, but there are bigger ones, I don’t expect to get realized in measurable time. As the single components can be done within a few months, I think a sponsor would be great (even if illusive). Anyway, who of you could take time for a few month (so student or freelancer?)?
Topics mentioned by you sound interesting. Nevertheless we have more then one 3d solution. Are you talking in the context of founding features to only one application? Or the result of codespring will be more general and available for other application too?
If you mentioned about infrastructure I’m extremely interested in finding server for 3d tile render. It produce quite interesting result. Application is ready from some time now. Only server and database is need.
To tell you the truth, I have no idea how to exactly take a choice on which project to sponsor or if we can improve all of them. I guess this will depend on the amount of money that we get offered. If you expect 40K Eur (that is a regular wage here in Germany) and think of 5 developers, we get total 200k Eur we require for a sprint
Any ideas, how we can boost the development, without getting some tools “lost” if we couldn’t sponsor all of them?
Concernig the infrastructure: Wouldn’t it be usefull, to write a general service, that supports different renders to get pluged in? Currently I see a lot of approaches to process/ship 3D raster tiles and everybody starts from scratch?
I agree with the general sentiment that funding development would be a great boost for 3D OSM - I’ve been working on OSM2World for more than three years now but there’s only so much I (and probably most other developers) can afford to do in our spare time. So I would gladly participate in a paid/sponsored development effort.
It would be useful to look for opportunities to work together more closely and share implementations of functionality which more than one of us wants. This may include code for
serving raster tiles
serving models for interactive clients
editor integration (JOSM)
and other applications consuming 3d models. Right now we do indeed duplicate a lot of effort here.
With Kendzi and myself this should even be quite easy (in theory ;)) - we are already using the same programming language and graphics API. It would require spending some time just on designing interfaces and refactoring, however, rather than adding new features. And I’m not sure whether we can find an agreement on all the relevant issues - the last time we discussed possibilities for cooperation, I got the impression we had some slightly different goals. It would still be very beneficial if more cooperation were possible, though.
As a side note, I’m currently working on the topic of integrating OSM with elevation data and hope to finish a first working solution in spring this year. Just to avoid more duplicated effort.