@timmehhhhhhh , totally understand your concerns. Microverse will continue to be relevant and applicable throughout the generational release of Unity 6. By the time we are able to consider this new system production ready we should also have a more clear picture of the roadmap and timing for splines integration. And we fully intend to get our key asset store developers on board as soon as possible so that they can continue to add value to Unity by building on top of the new system. That is key from our perspective.
We will definitely be able to bake terrains to meshes. This will also be the ideal way to target WebGL as it doesn’t support compute shaders. However, the meshes won’t be able to make use of the virtual texturing or tessellation.
Not sure how to answer the data storage question and may have to defer to @keesrijnen on that one.
In terms of Houdini, it will not likely be a full Houdini integration unless Houdini and Unity can work together on that across the entirety of Unity’s tooling and systems. That being said, you will be able to import heightmaps, mask, and pre-scattered assets from Houdini as point cloud files. We know that much.
Hi @logicalerror we know that holes are critical and intend to have that functionality by the general access, or production ready release of this system. It won’t be a full CSG system but we intend to support holes.
Loved the Roadmap video but the example of why you would use AI was simplistic and more time consuming than simply adding a rigid body controller.
I would love to see something more complex where AI could save me a lot of time.
i.e. creating a score script after certain actions happen and connecting it to my UI element that would display the score.
Very much welcome LOD feature. Would love to start using it soon, and design assets with that in mind. But also, I would love to see my super-efficient instanced low poly modules not creating crazy millions of triangles when they are rendered…
Loved the Platform roadmap. Especially XR. Let’s hope that this time after many years, it will be as advertised.
The improved multiplayer workflow sounds efficient and exciting to those of us working a lot with prototypes.
For Unity NXG, I am a bit scared about the “Simplicity” part. Every time simplicity was mentioned in the past, it meant shoehorning the exact same workflow in every aspect of the engine and that is not exactly simplicity imo. I think, simplicity means different things for each aspect of the engine. Each one serves different requirements and workflows. Animation, audio management, lighting, art asset workflows, post processing, etc.
But I would expect Post Processing to follow the exact same UX paradigm in both URP and HDRP. Today it does not. In order to add and manage an AO pass in URP I need to go to the renderer and in HDRP it is simply a component of post-production that can be added to the profile right in the editor. Or managing the sky etc. To me, that would mean simplicity. Having different ways to do the same thing in the same engine confuses people. I have yet to work with a team that does not think so.
Whether or not V1 supports that, I think the goal is to have the layer system be extensible and modular, so you could make a tree generator yourself that has whatever features you want out of it.
At this moment in time, we generate the entities of scattered objects in parallel jobs, then merge them all into the main entities World.
So these entities do not exist in the authoring scene, only in the entities world.
And thus: no, they are not placed in any sub-scene.
However, we plan to do work on scattering before release so I don’t know for sure what the final implementation will look like.
And especially of interest, how do we handle the situation where you want to stream them in and out as the player moves in the world. TBD on that but love to hear people’s thoughts on what might work for them.
Will the building tools like stamping, object placing etc be GPU accelerated? Watching a video it seems when u drag the object around it lags. As far as I remember MV is GPU accelerated and updates takes super fast so moving stuf around is smooth and not lagging.
Second thing regardng virutal textures, do You plan to have something like Runtime Virtual Textures, that u can sample in shader graph and for instance tint the grass based on the layer it is spawned on?
Do You plan any kind of migration function (nearing the release ofc ) for existing terrains (migrating textures, trees, grass etc, objects spawned / texturs spawned via existing Unity’s terrain tools).
Its not terrain related per se, but it is kinda world building topic. Do You plan at any point adding some kind of built in, easy to use, streaming solution for larger worlds ? (Large world can be 4k+ ). With all those tools You want to include in the next generation i think it would accelerate ppl to create more open world environments which require good streaming solutions for performance. Hope You have some card about world streaming dangling somewhere on “Trelo” roadmap ;).
Sounds problematic for multi-world simulations that require the physics colliders in several worlds. In those cases, the scattering tools could not be used and would need to be reimplanted.
We use a different setup than MV.
We divide the world into tiles and recompute the data flowing into the Layer Stack per tile.
So any update lag you see is probably just different tiles being updated.
The reason we divide the world into tiles is multiple.
We think it will eventually scale to much bigger world and it also makes it so you can author the world once (at high res) and bake it to multiple platforms of different capabilities.
MW also needs to do this eventually by creating multiple Terrains. So it isn’t that different in the end.
In our system, if you want similar type of update you can hide the tiles (aka parts of the world) you aren’t interested in so that the ones that are still visible get priority in updating.
It really depends how complex your Layer Stack is whether updates will be realtime or slower.
As for sampling from the virtual texture to tint your grass, that is already possible.
As for a world object streaming tool, we agree it is going to be needed, but at the moment my team is focussed on the terrain surface and making that perform well (which includes streaming its texture data). So the object streaming may come later, or from a different team, or must be implemented yourself (or via asset store) until we are ready to implement it.
It is a good point, something we will need to address as we scale this system up.
In its current form I don’t think you’d need to reimplement all of the scattering tools though. You’d need to add some functionality to do a pass over what was scattered and place whatever data you need into other worlds. That doesn’t seem very hard to do and our code is also all in C# package so you can always modify it for your needs.
I know everyone is treating this like a wishlist thread but the most important thing Unity can do is create a solid foundation for gamedev and asset store devs to build upon. The community will be able to iterate on these items faster than Unity can. Especially when it comes to tooling to create terrain. So I rather Unity not waste more months and years on this kind of stuff. They can always circle back and add it in later.
Something like streaming is best done by Unity. As the output of assets/tools should ideally be in a universal format and Unity is best position to come up with such a format.
I’m not sure if this is the right thread to ask: but is there any solution being made for the automatic billboard creation?
Would be very useful for larger worlds.
@nickdollahz , I would tend to agree with you and this is what we’re pushing for. We will be pushing for a streaming option but this is tough to do as a one size fits all feature and really comes down to gameplay and the hardware of your target platform. It will also need to work with the terrain, lighting, physics, and more which is why it’s taken longer than we’d like.
@Redrag0n , do you mean, like automatic LODs billboarding or Imposters? SpeedTree has this and we’re looking into longterm options along these lines but don’t have a precise solution for this yet.
Having a generic billboards and imposter on demand generators and tools would be very use full, the current existing one are only works specifically for SpeedTree isn’t?
) for existing terrains (migrating textures, trees, grass etc, objects spawned / texturs spawned via existing Unity’s terrain tools).