Artist here - Just a disclaimer that I can’t program nor am I experienced in C++ or any similar “backend” languages - best I can do is blueprints!
Hello! I’ve been trying to recompile the Cesium plugin for 5.3 preview, because of 5.3 volumetric pathtracer features - support for pathtraced clouds and similar volumetrics. I am in a timecrunch and sadly do not have the experience or time to properly learn and understand what I’m supposed to do within Visual Studio to be able to properly recompile the plugin.
Here is what I’ve done so far and the issues I have encountered:
As I understand it, there are significant changes in 5.3 that prevent the 5.2 version from just being “recompiled” and do require code changes (that I plainly do not understand or have the time to learn about) hence my plea for help.
Would anyone please be kind enough to walk me through, or point me in the right direction, as to how to make Cesium work on 5.3 preview? Am under a time crunch hence the rush as opposed to just waiting for the proper release.
I understand the time crunch, but we unfortunately don’t have the time, either, to urgently jump on building the plugin for every new preview release of Unreal Engine. Particularly given that our plugin is free and open source. Perhaps someone from the community with more C++ experience is interested in tackling it.
Based on a quick glance at your errors, they seem to be related to Cesium’s rather deep integration with UE for:
Point cloud attenutation, and
Occlusion culling
Neither of these are “essential” (depending on your use-case, of course), so the quickest path in the short term might be to try to disable these features rather than get them working in UE 5.3.
We have a PR that adds compatibility for 5.3 on Github that has been merged in preparation for the next release. If you’d like, you can grab the binaries from CI and try it out now! Otherwise, just hang tight for the October release
Id love to see cesium worjing with stable diffusion inside unreal engine
Imagine a simple overhead camera capturing not only terrain data, but textures
Stable diffusion coukd be trained to recognize satellite imagery
Bushes, rocks, roads, rivers, water, trees etc
And based on the location the landscape texture wil generate megascans foliage
It would be quite another level to cesium, brushify, MEGASCANS and Ai generated textures
