So, Nanite

"Unity needs to catch up!" is a ridiculous sentiment. Unity attempting to play catchup with AAA tech is one of the problems the engine has, not a solution to them.

 

Or probably, Unity should probably make some other feature, that devs have always wanted, that is unavailable in Unreal Engine. But, if Unity is already working on something nanite like, then I hope it turns out really good.

 

How much longer can one continue with 3rd party assets that have been corrupted by a single update? We call it Nanite, a DOTS hybrid renderer is introduced in the middle that completely undermines the development process. We say Lumen, there is not the slightest improvement in real-time dynamic lighting. (RTX was already developed by Nvidia, it would be ridiculous to count it. Because this system still has many shortcomings in Unity.)

We call it Virtual Texture because in a huge world the boot system has to be flawless. The current Virtual Texture system is in disgrace. Unfortunately, we have seen that there will be no significant development before Unity 2023. (Unity 2024 maybe.)

We rewrote the Terrain system for our open world game. (I won't even explain about it.)

We wrote the World Partition system in another company's game engine ourselves in Unity. (It's a small system, but it seems to best fit the needs of game developers.)

We had to solve the Floating Origin system in the most primitive way. It shouldn't have been too difficult to integrate it into the engine on-board.

I'm the senior developer of a studio focused on developing AAA-level games on the Windows platform. As we tried to write these technologies one by one with our own hands, we started to think that Unity didn't even deserve a Plus license.

What I want to point out here is that Unity continues to improve things. But now I want to say that these works are extremely slow compared to game developers. As for Nanite, I wish the only problem was that Unity didn't have Nanite. Because the point you missed is that, nanite is a technology that can be written by a normal developer, but many more features that need to work synchronously with Nanite need to be developed. These features are well understood by those who are deeply immersed in technical terms.

Also, I am sorry to everyone for my bad english.

I hope Unity makes things better...

 

if nanite and lumen were tools that could be developed by any small studio, it wouldn't have been such a big deal to the point of bringing in lots of HUGE game and movie studios to use unreal, would it?

And who makes money for unity? Small devs that use the free version or big studios sharing a % of their revenue when obligated to buy their license?

 

Ads. Ads makes money for Unity. Also "big studios" are not necessarily going to be making "big games" because the majority of companies pulling in enough to have to buy Plus/Pro/Enterprise licenses are usually targeting phones.

 

It doesn't work on animated characters/props. Also. Unreal needs a proper retargeting system like Unity has had for nearly a decade. Unreal needs to do some catchup with Unity. Having the animations tied strictly to the imported 3D file is so dawn of the millennia.

 

Unity does not get a share of profits. Are you some Untreal fanbois invading the forums to boost it?

 

so they don't get a % of it when a project/company makes over ~100k/year with a game made in unity? hm

Also, no. Wishing the engine I use to have better tech(nanite doesnt just make things easier, but it also looks better than LOD tech, due to models still existing at a distance instead of disappearing or becoming a png) doesn't mean I'm a fanboy(???) of another engine


And finally, obviously, unity has dozens of bugs. I know that, I find a new one almost every time I build a new version of my project, but bug fixing should not stop feature development. If it does, thats an issue unity needs to address in the very core of their project management/development process

 

ding ding ding

 

Well you obviously know little about Unity. Unity charges a yearly subscription price and does not take a percentage. Unreal has the wrong axis system, blows up imported models with carefully constructed hierarchies and if I have twenty characters with twenty walk/run animations I cannot just set it up once and retarget. I have to make animations for each file. Bogus. At 50 bucks an hour for the run/walk anims that is 20X the cost of the Unity method. It looks pretty and amateurs can get good lighting out of the box..of one kind. Animated objects do not get nanite..so what use is it except scenery. Yeah. I am pushing it but you kiddies are so tiring with your coming over here to proclaim some kind of overall superiority. Unreal could not have created the last 4 of 5 projects I worked on. It has a limited framework and an obscure API. Plus.. I ain't gonna recommend a company to any client that wants to open their books and peer in to get a cut of their piece of the pie. Especially one from a communist country known for industrial espionage.

 

What exactly is hanging up your game that is Unity-centric? Perhaps some more experienced devs can help you dig your way out or help you banish the spirit of resistance that seems to hold up your progress.

 

Works with VR now, that's a game changer

 

You lost any credibility on this line

 

Unity license costs per seat.
So if you have 100 people on your team using unity pro, that's $200 000 per year.
Unreal takes percent of revenue. So if you have 100 people, but earned nothing, it costs nothing.

The situation goes like this. If you're a lone wolf that somehow earned a million used with another flappy bird, that's still $2000 per year with unity. If you got ten million, hundred million (you wish), that's still $2000 per year.

In case of unreal, first $1 mil costs you nothing, but then you'll start paying more and more.
From $10 mil you'll pay about $450000, and from a hundred million you'd pay $4950000.

In reality any huge project with prediced large revenue will negotiate custom license with unreal engine and they'll probably pay six figures for it or more. Same deal applies to unity, as anyone with huge budget and huge team will also use "contact US" option and likely pay six figures for a custom enterprise option.

However, since you're not part of the realm of big boys, the difference in pricing for you means this:

* Unity has no strong financial reason to care if you succeed or not. Unity is interested in you having a huge bloated team, buying pro, buying every service under the sun, and paying through your nose till the end of time. It doesn't matter for unity if your game earns you nothing, as long as you pay your subscription.

* Unreal at a glance appears to care about your success, because the more you earn, the more they get. But their cutoff range for royalties (first million is free) means they expect you to fail and your project to die and that they do not care about tiny projects beyond using them as means to advertise their engine. Their strategy is to give tools to everybody in hopes that that will produce commercial successes with 0.001% chance, and get their cut from those. And that those successes will continue using their tools.

 

Auditing lots of projects to get 5% of little-to-nothing still costs a bunch of money.

In any case, no engine vendor makes much money off small projects. Given that, it'd make little sense to have barriers to entry at the entry level. Everyone wants the market to gain experience with their tools, because when studios with real budgets start up a significant factor in choosing what tools to use is what the decision makers and/or available workforce have existing experience in.

 

I think unity makes loads off money from small projects. From asset store

 

unity financial results info here
https://forum.unity.com/threads/unity-financial-results.1154549/#post-8574509

 

Again, Unity have positioned themselves to a place where they don't need their users to be successful (and actually it's even better if they're unsuccessful) in order to make money, which makes sense (I mean, their users are using... Unity!).

They make most of their money from ads, they are an ad company, so all they need from the userbase is to flood the mobile market with a billion S***ty ad-ridden f2p games, which has happened.

Apart from that, it's optimal that a ton of people struggle to make games for years while using a bunch of S***ty Unity services, instead of actually releasing stuff. Which means they need a lot of flashy features, implemented in a haste.

With all that in mind, it makes a lot of sense why we get a gazillion services (that don't really help you btw, collaborate anyone?), features that fall apart when you demand publishable levels of polish or reliability from them, and why a lot of focus has been put on ads and analytics (to the point of merging with such a company).

So SURE, let's have them half ass a nanite, that will be good for us all... right?

 

So what exactly is hanging up your game that Unity is responsible for and not you.? Some of us more experienced devs may be able to help you work around it Unless your goal is to simply be an acid arrow.

 

Wasn't seeking credibility with that line. It did precisely what it was mean to do. Hence your reply. Unreal ain't all that and a chicken sandwich. Unity is way more flexible and extensible.

 

It must be, so you can create features that are already built in into Unreal for years.

 

Unity has features that are missing from Unreal, and you shouldn't forget that.

Mecanim has already been brought up (easy effortless retargeting matters for a small studio, and Mecanim allows UMA with variable limb length), then there's a matter of asset postprocessor, for which our resident unreal prophet has been unable to rememebr an equivalent.

There's also general mess that you'd need to deal with if you hate visual scripting. I also have fond memories about glitches while working on converter. For example, making a "prefab" blueprint through code would open editor. Then the horrors of dealing with guts of FbxImport. Then minor stuff like that generating a skeletal mesh from code would crash the whole editor, creation of AR applications being a complete nightmare and so on. And none of this crap was ever documented anywhere. Meanwhile in unity api this sort of stuff was trivial.

Add to that huge install size. Unreal 5.1 is a 100 gigabytes install. Unity install can be as low as 5.

 

Who cares about the install size when nothing works.

Though granted C# is a huge selling point. Just wish we could get rid of mono.

 

Oh come on, there are still weekly new games on Steam made with Unity.
https://store.steampowered.com/curator/39750107-Games-Made-With-Unity/ (217 in Unity)
https://store.steampowered.com/curator/39869360-Games-made-with-Unreal-Engine/?appid=1040420 (42 in Unreal)

Not to mention mobile where Unity excels and (often) lower quality ones on Itch.io.

So much for nothing works

Yeah, when that happens, Unity may get a big boost. Nice to read recently that they intend to hire more people for that, but I do assume that won't change too much to the timeframe of 2-3 years since they have to support so many platforms...

 

If you need them. Otherwise they are bloat.

 

I have written hundreds of thousands of lines of code, integrated dozens of complex API's and all manner of dlls and it all worked with the caveat I did have to find some workarounds and create many of my own API's and classes. Your FUD and statements belies the users abilities and not the use of the application in a production environment.

 

We are really not going anywhere with this logic.

 

Not if you don't understand it. It is perfectly clear to me. If I am doing a UI centric AI app then what do I need nanite, or photorealistic rendering for ..or conversely in Unity why do I need LOD and HDRI? Simply bloat for the type of application.

 

Lol guys its Nanite topic not UE vs Unity . I think at this point both engines serves different type of devs and games, so there is no need to argue about it

 

I'm making a real game, with medium to large scenes, not tech demo. And workflow is a complete pain. Lightmapping, batching etc. Everything is more or less manual and super janky. World building is really lacking in unity.

Sure the constant regression is also a big problem but i would say above is a bigger problem actually.

 

That's just straight up trolling at this point. Or lying.

If nothing worked, the engine would've been long dead and no new project would be released with it. Which is not the case. It is still quite useful and quite usable.

In this case the question is why you're still here. I'm certain you know where the door is. Just don't forget how constant engine switching turned out for Duke Nukem Forever.

 

Can someone in layman's terms explain this to me? Thanks.

 

Its means there is a bigger chance than not that the firm will go bankrupt.

 

Im offcourse here for the run, but next game might be in unreal. Though c# is a much better domain language than Cpp

 

For comparison, Twitter is currently also at 50% according to them.
https://www.macroaxis.com/invest/ratio/TWR.MU/Probability-Of-Bankruptcy

 

Unity will not bankrupt.
They got to big user base for this.

But they might be bought or they might reduce the scope.

 

By this logic every engine is bloated.

 

At the moment, it is true. Every widely used engine is bloated, and non-bloated engines would be inhouse ones (axiom verge?) based on libsdl.

Ideally, you should be able to disable subsystems completely if you aren't using them.

Qt used to use this approach. You had QtCore module, everything on top of it was optional and could be disabled.

 

Well, Unity has the package system for that purpose. E.g. neither ECS not URP or HDRP which are all massive things, are in by default.

 

Well, but last time I checked you can't disable audiosystem or physics, for example.

 

Not an expert, but bankruptcy can mean a lot of things. I.e. if twitter went bankrupt, it would likely result in a corporate restructure, reevaluation debt etc., not closing its doors.

 

https://docs.unity3d.com/Manual/class-AudioManager.html

You're right, you can't "disable" physics (it will always be present in your project as of today), but you can turn it off by turning off the whole collision matrix (just in case) and untick this:

And of course, never call

Physics.Simulate

.

(ps: this isn't really an argument, just providing extra info on the subject)

 

I would say unity is using physics system for some internal workings.
This is why you probably use it even when you don't use it explicitly.

For example Volumes are using physics colliders.

 

The question isn't "which is better?", but rather, "which is considered better (by more people)?" The former is a subjective question, and the latter is an objective question. The answer to the latter question may not align with one's own opinion, so it is often avoided, ignored, or dismissed.

The reason this "war" will never end is because it's waged on a subjective question rather than an objective one.

 

Lots of unreal fanboys who thinks a game engine is just nanite and lumen, lots of unity user who install the last alpha version from unity and then cry because it has bugs, i'm impress at how common that is.

Right now there are more games on steam made with unity but that will change soon because unreal 5 is VERY popular, the problem is that most of those games will be trash, the same as unity games on steam (in fact it already started)

...And for the guy or girl who said that size of installation does not matter, it matters, even for today standards, i remember trying a demo from unreal 5, you can walk the entire demo in less than a minute and i think it was more than 100gb, that is completely useless, i would prefer to recreate that on godot (and i do not like godot)

Ads are Optional... so i think your theory is not exactly correct.

I think Unreal right now is better if you have a pro team full of artist who knows how to make amazing realistic graphics (because is not just nanite and lumen and keep in mind that a 587483gb game is useless for the majority of gamers and i'm not even talking about the stability of the game itself (or the game engine) i have a couple of unreal games here and i'm pretty familiar with the "crash reporter" which is much better know that the one from games made with unity, so... ¯\(0_o)/¯

 

Why are you so misinformed?

Source: https://docs.unrealengine.com/5.0/en-US/nanite-virtualized-geometry-in-unreal-engine/

Nanite is an obvious upgrade to current mesh optimization just like how Virtual Textures was when it appeared.
There is no downside even with low poly mesh. You don't need to have super high poly mesh to benefit from Nanite.
The demo's size is because of the stress test they put into for showcasing its power. There is no need to use mesh for these chain links.

And you dare doubted Lumen's value?
The number of hours wasted with just dealing with baked GI issues is ridiculous at this point, its very invaluable to both small and large companies.

 

Its almost 2023, and we still cannot change Animator target transition. What to say about anything else.

Upd: stock price says its all
From 120 to 36 in a year

 

What I also noticed is that some indie games were created with UE, but also have the same problems as some Unity indies games. Such as poor performance, lots of bugs, crashes...
If you then compare such games, especially games with similar game mechanics, there are big differences. E.g. base building/automation games.
If you take Satisfactory as a basis here, where you can build and automate a lot in a large world, so that the limit of the maximum objects of unreal objects can be reached.
Other games, on the other hand, already have performance problems and instabilities with a small build quantity. As an example I can name Hydronieers or Astro colonies, the performance has improved, but they are far behind Satisfactory or similar games that were created in Unity (like dyson sphere program).

 

You can disable almost all Unity internal systems in Package Manager (Packages: Built-in). And you can even disable all Physics (and it won't be included in build).

Of course if you disable for example Physics Package you must disable also all packages which depend on it, so there won't be compilation errors

 

Well..That's just dandy for archviz. However the vaunted immersion breaks with animated characters/props/scenery that can not get the same treatment.