DataDiVR Backend

This is a flask server that provides the network data to the UnrealEngine DataDiVR VR Clients

DataDiVR VR-Module

Please find the build of the newest DataDiVR VR Module here. To run the Module download the folder, extract it and execute the .exe file.

INSTALLATION

1. run backend

   • install python 3.9 (for mac we recommend using brew and installing it like this: brew install python@3.9 if you use other ways you might need to change the mac_buildandrun.sh file accordingly)
   • windows: run Buildandrun.ps1 in console
   • linux: run linux_buildandrun.sh in console
   • mac: run mac_buildandrun.sh in console

   If all dependencies installed correctly, the console should show
   The server is now running at 127.0.0.1:5000

   As on mac port 5000 is already occupied by the systems control center, on mac the server will run on port 3000 (instead of 5000).

2. upload data and create new project

   • choose which file format you would like to use (CSV tables or JSON files)
   • open a browser (Chrome or Firefox)
     • for CSV file upload go to 127.0.0.1:5000/upload / 127.0.0.1:3000/upload(mac)
     • for JSON file upload go to 127.0.0.1:5000/uploadJSON / 127.0.0.1:3000/uploadJSON(mac)
   • make sure "create new project" is checked and assign a name to your project
   • choose one of the provided examples files (available via download on the uploader webpages) or use your own
   • click upload

   After a success message was shown, the uploader has now created a new folder in "static/projects/yourprojectname" containing all the data in the DataDiVR format.

3. use the WebGL preview to have a look at your project without having to use VR hardware

   • go to 127.0.0.1:5000/preview / 127.0.0.1:3000/preview(mac)
   • select your project from the dropdown

4. run the VR you need:

   • a VR ready windows computer
   • a SteamVR compatible headset
   • SteamVR installed
   • download the DataDiVR executable
   • open "DataDiVR/Colab/Content/data/config.txt and make sure it contains the adress the DataDiVR backend is running at
   • run the backend
   • run DataDiVR.exe
   • enter a username (optional)
   • choose robot if you want to run in desktop mode (good idea for a dedicated server as it is more performant) OR VR to run in vr mode
   • the first player is the host so choose "HOST SESSION"
   • the following players choose "JOIN SESSION"

DOCUMENTATION

Once the flask server is running, go to 127.0.0.1:5000/doku / 127.0.0.1:3000/doku(mac) to learn more about the DataDiVR framework

DataDiVR Dataformat

The DataDiVR_WebApp acts as a multiplayer gameserver for one or more VR clients. Its purpose is to serve the connected players with big network datasets - as quickly as possible. That is the reason why most properties are stored (and transmitted over the network) as images.

Every folder in "static/projects/ contains 3 JSON files (check out the file dataframeTemplate.json for the exact structure) as well as 5 subfolders containing textures

• static/projects/projectname/
  • nodes.json
  • links.json
  • pfile.json
  • pdata.json
  • layouts
    • layout01XYZ.bmp
    • layout02XYZ.bmp
  • layoutsl
    • layoutl01XYZ.bmp
    • layoutl02XYZ.bmp
  • layoutsRGB
    • layout01RGB.png
    • layout02RGB.png
  • links
    • links.bmp
  • linksRGB
    • linksRGB.png

layouts + layoutsl -> Node Positions

this needs a little explaining: Think of a texture as a dataset of the following format: [[R,G,B],[R,G,B],[R,G,B],..] where every [R,G,B] is a pixel. This can be used to store a location (X->R Y->G Z->B) per pixel. Because a .bmp only has 8 bit depth we need a second texture to get a resolution of 65536 per axis. this is where "layoutsl" comes into play. NOTE: node positions need to be in a 0 - 1 range (!), the conversion works like this:

floor(x _ 256) -> layouts floor(x _ 65536 % 256) -> layoutsl

This means, that the available space is not unlimited, so when nodes are closer than 1/65536 units they will snap together.

HOW TO MAKE YOUR OWN USER INTERFACE

The User Interfaces for the DataDiVR are realized with html and js and are rendered in the UnrealEngine in-game webbrowser, which is Chromium. Data is passed between the flask server and the html clients in JSON format. The html pages also act as a middleman between the UnrealEngine VR Module and the flask server. Here is a series of examples that explain in detail how to create your own user interfaces. (you have to run the flask server locally to see those pages)

go to 127.0.0.1:5000/doku / 127.0.0.1:3000/doku(mac)