So i somewhat recently started my journey here on Unreal engine
and i stumbled across a video about World Composition
vs World Partition level streaming also seems to be a third way to do it or part of World partition not 100% on that
World Composition seems to be better than world partition at least in the way that it works however its deprecated as it lets me know
Is there any reason not to use this method? seems easier to do and possibly much more performant
if somebody could give me the basics of what each one does so i know the whys of choosing one vs the other and potential performance implications
I’m pretty sure World Composition is the redacted version of World Partition. Both of them basically use level streaming to operate. They are supposed to work autonomously, and are both focused around vast open worlds.
But, you can write your own system with level stream nodes ( either volumes or blueprint ) if you have different ideas about how you want the streaming to happen. More control, smaller worlds, streaming in Z etc…
So no big downfalls if used correctly?
I watched a video on World Composition and it seems to be an easier workflow method
World Composition:
i can have the main level be say background mountains and things
then its children maps would contain things that you don’t need to see unless your fairly close easily established with the layers
so an entire villages buildings could be one group and the props around the buildings another level
This seems fairly ideal compared to my current understanding of world partition
World Partition:
The landscape or rather the same chunk sizes that the landscape uses determines what loads so not um how to say per asset
you end up with it looking like Minecraft style with Partition loading where your far more likely to see these chunks not loaded than if you use Composition where all relevant terrain can be chosen to be loaded and the assts or buildings etc can then be loaded when your near enough making it super easy to have what i like to call far away meshes ex a cube representing a sky scrapper and when your within actual seeing distances it can swap to the more detailed meshes and load the props
i have to continue to do my research on either or both clearly currently i just have a persistent level which works out okay for me with the beefy pc but to say the least one of my testers thought i had some kind of motion blur on
spoilers they were lagging from loading the size of the world
which isint anything crazy crazy i don’t think something like 4000x4000 or 4km2
hopefully this makes the slightest bit of sense
my main concern is if its a worse idea to use Composition than learning how to do level streaming and world partition correctly as its a not deprecated and presumably newer
games like skyrim and gta v seem to use composition from what i can see and understand currently
This whole posts secret agenda is to learn more vocabulary and how it relates to further my searches and understanding of how it all fits together
