Use UTM to standardize projection about blosm

[vvoovv]

I think Blender is a 3D modelling software, not a specialized GIS software. So here are my conclusions taken into account that point.

1. The key task of geographical importers is to load data with real dimensions without distortions.
2. I don't think it's a good idea to overload Blender users with very specific GIS terms like projections, SRID, EPSG codes, etc.

The current solution with Transverse Mercator works quite well to solve those problems. So I'd prefer to keep it as is.

[domlysz]

Thanks for your answer

I understand your opinion, especially keeping the OSM importer simple as most as possible is important considering the large audience that may interest.

For my part, I'm not afraid to introduce GIS terms because its how geographics data are handle at all and I can't build importer for specifics formats like shp or geotiff without talking about georeferencing or projection.

However, I'm not really in favor to introduce reprojection capabilities to Blender, even if I finally come to this to improve my tool and offers new interesting functions. For me it's an help for solving some situation not a finality and I really encourage users to preprocess their data beforehand, in a dedicated software.

The important thing is not reprojection but georeferencing. Some people ask for compatibility between addons which handle georeferencing informations and I think it's significant. But to achieve this keeping track of all georeferencing infos is essential : scene origin geolocation and map projection that all, not a big deal.

I can hack my tool like dxf importer to handle the special case of blender-geo transverse mercator however it will be better if we can converge to a more standard way, and need your cooperation here.

I don't think this modification have a huge impact for end-user of osm importer:

• you don't need to bother user about reprojection, SRID or EPSG, just add this information in scene propeties, most of user will don't care about this but some others can rely on it.
• The scale error within each zone of UTM distortion remains less than 0.1 percent. I think its very acceptable and can't imagine any user complains about this.

Hope you reevaluate your position !

Regards

[vvoovv]

The current solution with Transverse Mercator works perfectly for me and other users of the blender-geo add-on. It perfectly solves the problems I mentioned in my previous message:

1. negligible distortion
2. simple to use

So I don't see the need to use other projections.

You may assume in your code that if the Blender scene has custom attributes longitude and latitude and no specific attribute about projection, then the Blender scene is in the Transverse Mercator with the center in longitude and latitude.

[domlysz]

ok no problem understand your point