Dealing with 3D

[sjib]

In TEKSI we have started to implement geometries as 3D geometries to support the collection not only x/y but also z-values.

The INTERLIS data models all have a 3D extension, but it is not widely used yet:

• e.g. VSA-DSS 2015 and VSA-DSS 2015 3D
• e.g. VSA-DSS 2020 and VSA-DSS 2020 3D

A reason why INTERLIS 3D is not used up to now is that we currently have a "problem" in INTERLIS:
We can create e.g. detail_geometry3d, but will not have 3D information in most cases for a whole community right now. And if we do not have that, then the validation of INTERLIS VSA-DSS 2020 3D fails because different to the database postgres, INTERLIS can not have NULL for the z value.
So while x,y, NULL is valid in postgres, this is not a valid geometry in INTERLIS.

So we have to think what to do, if we have 3D geometries in qgep and want to be able to move those to tww. For the begin / and of a POLYLINE we often have a separate attribute like "Kote", but we do not have separate attribute for intermediatary points in a "Verlauf / Progression".

Possible approaches:

• Make the INTERLIS - INTERLIS transfer with the 2.5 D version and at the end update "progression3d" manually after the import
• Integrate the 3D Version on export of qgep and on import of tww and convince the VSA to support also INTERLIS - INTERLIS 3D in 2015 - 2020. This would need a convention on how to handle z= NULL for the transfer so we do not throw a bunch of errors on all 3D geometries
• other ideas

[sjib]

some other thoughts on 3D https://github.com/orgs/teksi/discussions/104