I recently attempted some 3D mapping of churches ans castles I know well. Now, confronted with a half dome such as this feature, how would you describe it? In the same way, I tagged a round tower roof as “pyramidal” although this isn’t in my opinion an appropriate way to describe the conical nature of this roof.
I resorted to these compromises based upon the shape list of the simple 3D page. Do some of you have improvement suggestions?
roof:shape=dome makes sense here. For reference, this building part of a train station is also tagged as
roof:shape=dome and is shaped like a half-circle with an elongated side. It looks OK in F4Map, albeit with some z-fighting.
I used a roof:shape=dome for this church. I think successfully (F4map).
For common patterns of churches and castles it may be worth having a wiki page with some basic examples. For instance this is a very common style of medieval parish church in England. I’ve had rather less success with representing apses with a set of skillions (church in Zakopane), but if someone has evolved a successful approach then it’s likely that the approach can be copied.
Thanks a lot for your replies. This is helpful.
It seems that there is also a solution to get a good approximation using skilion and the JOSM building tools.
The issue is mostly solved using only the simple 3D standard. Here is a look at the Aigle castle. (I know there are still some issues with the main rectangle tower, but it shouldn’t be very difficult to tag it properly)
edit: the advices of @Tordanik on the old forum were very useful.
After about one week as a beginner in 3 D mapping, I have summarized my challenges an solution in a small diary post.
The spliting hack is too tedious. OSM-4D/Roof table - OpenStreetMap Wiki would be more usable, although wishful. Outside of this “standard” in
roof:shape=, the accepted
=onion (supported by https://wiki.f4map.com/render ) at 2769 instances is not that much more common than the 756
=cone. OpenStreetMap Taginfo
Yes it is too tedious but it seems to be the only one feasible within the the simple 3D building standard. As I wrote in the post, I thing the standard should be revisited.