Hello,

I'm working on terrain creation that would be split for composition in unreal with PDG and houdini-engine-v2. I followed that awesome tutorial: https://www.sidefx.com/tutorials/procedural-desert/ [www.sidefx.com]

However I had few PDG setup problems and I would like to understand a bit more about tags:

Here is my setup:
SOP HDA to create terrain (terrain_setup.hda)
-- One output (I would expect SOP)
SOP HDA to setup terrain (terrain_split.hda)
-- One SOP input to get the height field
-- One Output (I would expect SOP)

TOP node
-- hda processor from "terrain_setup.hda"
---- Create output with outputtag="mytag"
-- wedge to create attribute tile_NP
-- hda processor from "terrain_split.hda"
---- Reads input with "mytag"
---- Create output with "file/terrainsplit"

If my output tags (in asset output) are "file/geo" for all my hdaprocessors, I have the following error:

ERROR:root:Traceback (most recent call last): File "C:/Users/pouke/AppData/Local/Temp/houdini_temp/generate_terrain/pdgtemp/8128/scripts/hdaprocessor.py", line 876, in main() File "C:/Users/pouke/AppData/Local/Temp/houdini_temp/generate_terrain/pdgtemp/8128/scripts/hdaprocessor.py", line 872, in main run_job_mode() File "C:/Users/pouke/AppData/Local/Temp/houdini_temp/generate_terrain/pdgtemp/8128/scripts/hdaprocessor.py", line 736, in run_job_mode result = hda_service.run_job(work_item, temp_dir) File "C:/Users/pouke/AppData/Local/Temp/houdini_temp/generate_terrain/pdgtemp/8128/scripts/hdaprocessor.py", line 435, in run_job result = self._setup_scene(work_item, temp_dir) File "C:/Users/pouke/AppData/Local/Temp/houdini_temp/generate_terrain/pdgtemp/8128/scripts/hdaprocessor.py", line 408, in _setup_scene self._setup_file_inputs(work_item, hda) File "C:/Users/pouke/AppData/Local/Temp/houdini_temp/generate_terrain/pdgtemp/8128/scripts/hdaprocessor.py", line 346, in _setup_file_inputs self.installed_hdas[hda]['hda_node'].setInput(i, filenode) File "C:/PROGRA~1/SIDEEF~1/HOUDIN~1.408/houdini/python2.7libs\hou.py", line 13211, in setInput return _hou.Node_setInput(self, input_index, item_to_become_input, output_index) InvalidInput: Invalid input.

From what I understand PDG reader can't make difference between different outputs created by different hda processors.

Someone could explain me how tags are working in PDG, why is it set to "file/geo" by default and where tag are saved?

Thanks

For those who are stuck with tags and tops I fixed the problem the following way:

In the first hda processor, Set the output tag with anything. (eg: file/geo/initalterrain)
In the second hda processor, Set the input tag with the same value. (eg: file/geo/initalterrain)

If you want two inputs in an hda processor you can partition two inputs with the same tag and use the following tag in the hda processor requiering 2 tags.