You can do that by adding a global mesh... Tricky but doable.
For example, import the Phobos mesh into your 3D program of choice, model your base, export again.
The same can be done for Earth.
In theory yes - practically not. Phobos is a small potato of the size of
Manhattan New York. Earth is HUGE.
A single global mesh would be badly inefficient for larger planets and not needed for the gas giants, which could need their own improvements, IMHO (More cloud layers, own scattering models and procedural volumetric clouds)
I agree that the current base system must go.
I agree. The current one is fairly user friendly, though.
Maybe a community should make a prototype application for a solar system editor first and use that together with Martin and the visualization crowd for getting a better successor, that is also still user-friendly, by the means of being released together with a suitable editor.
Would at least help describing, what we want.
I also want to be able to create buildings, that have a DLL, but are no vessels. A lot of the vessel API is simply wasted there, while other features are badly missing (especially regarding level of detail management)
What we must be able to define are "landing places" of two types: pads and runways. And for those we need to be able to define the altitude.
The the terrain can be drawn procedurally or using meshes. If not terrain is used, set all altitudes to zero (legacy model).
It would also make sense to define things like taxiways and aprons. Those could also make sense in the simulation eventually (ground friction, damage modelling).
All scenery is scenery, not instrument navigation, I agree with that.
Yes, both is related since navigation beacons are often also scenery objects, but it is not the same layer of information.
I think that, for flying, less is more. If you want to fly to a special landmark, you can still look into the map for the coordinates and set up waypoints.
