3d arcs

Is it possible? to create 3d arcs and i mean not like that with the round roof:shape but more like bended line,bar like these in the picture

Or if it’s not an easy way what would be the easiest of the hard ways? :slight_smile:

Create an arc in a 3D editor. I recommend Blender.

We could extend the OSM-Tagging but I thing, it would be the wrong way.
Complex 3D models should be 3D files, inserted into the OSM world by the 3d renderers.

We once had a place to place model files:
https://wiki.openstreetmap.org/wiki/OpenBuildingModels
It’s offline.

And it’s time to set up a new one!
Inclunding a tagging proposal how to reference to the model file,
how to position, rotate and scale; what buildings are replaced by the model etc.
(Who has done proposals already and likes to help?)

-karlos-

vvoovv -

thanks but after i created it in 3d editor then what?,i want to upload it to a F4map

and OSM import for Blender 3d is to import OSM files to Blender not viceversa isnt’it?

Thanks karlos - i’m going with you.It would be very nice if we would have that server.By the way why OpenBuildingsModel is down?

I would like to help but i don’t know how.Maybe if it would be a way for everyone to help somehow,then maybe would be getting somewhere

I think, there is no upoad to F4map. You may ask them.
If we have a model server to upload also your model, they may use it.

OpenBuildingsModel is down because the hardware burned away, all students are gone
and the professor doesn’t have time/priority to seache for the existing backup files.

That will not stop us to set up a new server (at redundant places)
and may motivate others to upload new or find old models

Yeah i know there is no upload directly to F4map,i was referring to uploading it through OSM

such a pity,hmmm why all the students abandon it?!

I’m happy to here that…i’m waiting and Good Luck!..if I could help somehow…

I made specification for extending existing S3DB with such elements. It could work. I had as supervisor an very good IT student who has written his master thesis on this subject.

BUT:

  1. Specification has his complexity. I don´t konow if the community would accept it.
  2. Implementation should be done in f4 maps as well.
  3. Implementation in JOSM as PlugIn should be done.

Even if we would have point 3 I´m not sure if f4maps agree with that.
We should talk about it. I miss it still in 3D modeling of historic cathedrals :wink: .

Marek, is your specification published somewhere?

Also note that there are currently much more data consumers that support Simple 3D Buildings. See the one in my signature :).

The specification must be simple enough in order to be widely accepted.

Because we have to thing this in a bigger context, it extends this topic “3d arcs”
and I do a new one.

Dear Vvoovv,
I start with review of this specification and provide it on an separately wiki page.
We can see it as a kind of cooperative work on 3D description for the future I guess.

You are right. Simplicity is everything. Unfortunaltely reagding 3D the mapper should be able to spatial thinking and understanding of specification.

In advance about 3D arcs: I see this as some different cases:

  1. Arc openings in walls and another 3D elements.
  2. Free 3D profiles along spatial paths.

For case 1 the definition could be more easy than for the case 2.

In general my dream is to bring the Blender and 3D OSM community together.

I’m glad for carrying over this topic,thank you!

I would like to help also,but unfortunately i don’t think i can,i’m not a programmer but if i can help through other ways…

and forgive me,but i don’t understand;since we have those 3d shapes which we have already, why is so hard to implement more complex 3d shapes.
for example we have that round roof shape,so we have the 3d arc already but just on top part of the shape,why can’t be created also the bottom part…maybe develop it with somekind of substractions(booleans)…one shape subtract the other one…i don’t know,it was just an idea

Sure, this would be nice. I have for this reason the tag: thickness=*
This should be measured from the top surface to down, so still the height=* would be the highest height of the surface…

yeah, That’s what I think .