How is it that Unity has not approached this man yet?

How's that going so far? SRP hasn't done so well, even though it isn't bleeding tech.
And like I said, a small team working on tech like nanite, isn't going to effect editor performance, UI, animations, and core tools in general.

Doesn't have to be nanite, this applies to tech in general. It's unreasonable to just stop working on new or improved tech because the editor isn't perfect.

 

Gee, I wonder why it isn't going well? I wonder how thinly spread Unity is right now with constant acquisitions and a focus on a whole bunch of things that have eventually been left languishing? Do you think that putting a small team together, largely of graphics programmers, is going to somehow help things like URP reach feature parity with built-in? Or do you think that that's going to draw resources away from that already messy problem?

Unity has been suffering from "looks nice on the back of the box" acquisitions and additions since the Unity 5 cycle started. This has been the subject of a lot of complaints about the direction the engine has been taking. I even made a post about it in another thread!

All of these things are much more important than a fancy new toy. This kind of thread happens literally every time another engine gets a bleeding edge feature.

 

Unreal goes for Graphics, Unity goes for (game) mechanics.

I think games need to evolve, not everyone want's to play a movie, I prefer cool gameplay over cool graphics at any time.
Take Elden Ring for example. It's not the best looking game of the year but people actually enjoy playing it. Destructable Environment, interesting AI and Combat, cool Loot and no Microtransactions.

Was Minecraft about graphics? What about Terraria? Looking over the actually good indie games, they all have either a good story or good gameplay or both. None of the top ones I know of, sold because of graphics.

It's almost the other way around, I see a ton of high graphics walking simulators that fail at everything else. Asset flips. No story, no gameplay, no game.

 

Epic is doing the same thing -- bleeding tech (more work actually, check out how many packages they have), acquisitions left and right, and a major game on their hands. a blown full In engine modeling system, fluid simulation, procedural geometry, Realtime GI, an entirely new physics solution, tons of animation tools, an entirely new audio solution, and a bunch of other features.
Yet one is falling from a cliff, and the other is climbing.

I agree that SRP is an issue, but I don't believe it couldn't have been done well. I don't believe it's simply because they're spread thin. They didn't remove enlighten and bring it back because they're spread thin.
Many promises made and never met, timelines never met.

I'm not asking for nanite, my issue is with the idea they have other work they should focus on.
URP's snail development pace isn't simply because they don't have enough developers.
Point light damn shadows only added in 2021 isn't because they're spread thin.

Maybe it's management, maybe it's something else. But I don't believe it's because they don't have enough developers.

 

Games aren't just about graphics, absolutely. In fact I'd argue most gamers care for gameplay more than graphics.
Especially PC gamers.

Thing is, unreal 5 isn't just graphics. It brought new animation tools and improvements. New physics solution, an entire new audio solution, and a bunch of other features.

My point is they aren't just focusing on one thing, they have a ton of work to go through, same as unity.

Also personally I always found dark souls & elden ring graphics beautiful, I've seen a lot of people complain about it.
I suppose some mean it doesn't look realistic.

 

Because Epic's focus is bleeding edge tech! I have literally said this like ten times.

 

What? My point is epic isn't suffering as unity. It's not like they have developers that are specialized at bleeding edge tech.
that is to say unity's issue isn't that they're spread thin, which you claimed.

 

They literally do have that! Just like Id does! They have entire groups of developers whose focus is on new technologies! This is a thing they actually do!

 

Everything has viable use cases. You need to consider why we're no longer using direct voxel rendering.

That's actually a red flag.
Seems like Blueprints all over again. Where a company gets an "amazing idea" and then becomes so fascinated with it that it blows inordinate amount of resources on it.

 

I'll take your word for it. And assume unity has none of that.
Although I still believe unity's issues aren't that they're spread thin. There's nothing bleeding edge about basic features that existed for the last 10 years. Point light shadows, TAA, SSR, LOD fading, the list goes on and on.
to top it off animation seems to be completely halted. I'll give unity the benefit here and say that they moved to dots (even though they have nothing to show for it so far and animation is coming quite a bit after 1.0)

Unreal is also doing things left and right that aren't bleeding edge, UE5 comes with a lot of just that.

 

Unity spent so long adopting Back of the Box features that they had to announce that they were going to put a focus on fixing up the engine for an entire version release.

 

Unity In a way only releases one major version per year. Year.1 releases are mostly just bug fixes and small improvements. Almost all new render pipeline features are packed into .2 (check out 2021.1 and 2022.1)

 

Could be, only time will tell.
Hopefully time doesn't do what it did with SRP

 

As software developers who have learned text-based code from the very beginning, we are surely scoffing at blueprints, but for artists, designers and newcomers they are something quite useful.

 

The problem isn't the existence of blueprints, it's trying to do anything outside of the scope of blueprints. Extending UE4 compared to extending Unity is an absolute nightmare and that's a major problem when lots of games require custom solutions.

 

I agree with some points raised here that unity should focus on optimizations and general improvements (which it has already been doing) but to dismiss this technology as useless or not important is not smart to do.

I think Nanite technology is amazing, it allows you to use models with a high amount of polygons, and maintain good quality from near and far. regardless if your model is high poly or low poly, the benefit is that you don't need to use LODs, you don't need to remake the same object in several quality layers. You create just one, and the engine itself takes care of optimizing the mesh at runtime.

I think of this as something fundamental that would be great if it were integrated into DOTS or ECS. DOTS allows a massive world full of objects, with Nanite, NanoTech or any other similar tool, it would allow us to create a world rich in detail, without fading LODs or impostors. The artist makes the model, imports it into unity, and voilà, the world is created and optimized, without wasting time... Even for mobile this would be great.

But I could also be wrong, some other companies have already done optimization research like this, Ubisoft did it in 2018... the nanite idea is not new, but unreal was the first to be able to implement this whole pipeline without it falling apart. maybe this is the reason why nanite now exists, someone at epic had an epiphany and found a way to put everything together into one thing. i don't know...

This is the technology that other companies were craving and now have access to through Unreal, Unity should provide something similar or at least close to what undeal provides now to maintain good competition.

 

It was less an epiphany and more 10+ years of work on the part of the person who helped implement it into the engine.

https://twitter.com/briankaris/status/1260590413003362305

 

Exactly and also don't forget that this does not only work when you have a handful of ultra-high-polycount mesh instances. It also applies to having hundreds of mid-res meshes like it's common when designing open-world games. That way even small studios or indies can easily end up with a way too high polygon count if they do not put careful effort into LOD balancing.
Nanite might only make previously impossible, possible for AAA studios - but it does make work itself easier for smaller studios!

That said, I agree of course that Unity's LOD system which requires multiple game objects and a separate controller is quite an annoyance. Why not build LOD directly into mesh renderers?

Fear that's the problem if you are based on C++ :/ But yeah, an alternative might be a good idea, but it seems to work out for UE well so far.

 

Not really. O3DE's Gem system allows for much easier and better extensibility and is done with C++.

 

I understand that it is more accessible to non-programmers, but severe focus on blueprints means that I cannot utilize my main strength to the fullest. Meaning while they're having fun, I need to tie my hands behind back, stand on my head and code with left foot to do anything.

Not really. C++ allows plugin frameworks and dynamic libraries. 3dsmax and maya back in the day offered a very clean plugin api for extensions, which supported multiple different compilers even. Pure C++, no source code access.

Now, C++ does not have metadata, but this is easily solved by introducing a few mandatory functions.

So the problem lies really with unreal. They're not even using normal C++ in the first place. It is a dialect.

 

We don't need multiple threads to talk about Nanite, please use the existing one. (it has better title).
Closed

https://forum.unity.com/threads/so-nanite.889714/