A Unity ID allows you to buy and/or subscribe to Unity products and services, shop in the Asset Store and participate
in the Unity community.
Discussion in 'General Discussion' started by LIVENDA_LABS, Nov 12, 2019.
And more ...
Sorry, but what BS feeding.
Keep telling yourself that. Meanwhile I'm working for a studio making a game and not having the shiny toys that come with the competition hasn't held us back. We're even taking advantage of the latest rendering tech that someone claimed is only being used in very limited ways.
It's called first hand bug fixing and feedback, large companies don't just throw away multiple iterations for the fun of it. Modern engines are large, intricate and used in unfathomable amounts of scenarios so multi-bed / multi-stage testing is never a bad idea.
Sure I like many prefer Unity for its ease of use but the iteration process with a game engine used to actively develop titles IME sure is different, generally more practical and better thought out holistically. Whether one wants to use it or not is an entirely different matter.
What game are y'all honestly expecting Unity to make?
Because the answer always seems to end up being "the game I'm making because that covers my use cases."
Unity is a kitchen sink game engine that has to cover a lot of use cases and doesn't have its roots in first/third person shooter design like UE has historically.
Be nice if Unity fixed up Vulkan for Quest and optimised it. Also I'd love that dots netcode sample and lots of docs or sweet samples. I would like that stuff for xmas.
Vulkan is kinda broken everywhere, I swear it was in better shape like 16 months ago. But at some point it became slower than gles3.0 or flat out not working properly (on various Android devices).
RTX lightprobes??? gimme gimme gimme.
Cuphead 2.0 FTW
The problem is that they hire hyper specialist, but lack the team culture, these people worked on game sure, but studio culture are different and transcend individual, a lot of time, big shoot going to another studio usually doesn't translate into success, mostly it's when team move together that you can see continuity.
There is the workspace architecture, how informations flow, how issue are deal with, how responsibility is distributed and where are the bottlenecks. Organisation matter. You can't just have a collection of individual, you need a process that encompass and integrate everything.
Making a game stress all of that, you can tell unity have issue because it feel fragmented, a lot of silver bullet solution is typical of a fragmented organization. Like many people have pointed that the shadergraph and vfx graph aren't using similar metaphor, while doing almost similar thing, in fact same node are implemented using differing assumption.
You can see that in they are only addressing now some stuff that were stuff evident from the start, like issue with big project a lot of their copenhagen presentation addresses. I have always said they should do a big open world type of project, because it was obvious some decision wouldn't scale, and have proven to not scale.They are still making mistake, in the face of what everyone tells them with the URP and shadergraph, and needed hot fix like custom node.
Unreal isn't perfect, but at least it feels complete and integrated to make game, it feels like they understand that they need to offer a gold buffer instead of silver bullet. You understand there is a conscious workflow it was design for.
And you can tell fortnite really had an impact on efficiency as it's kind of the perfect "complete" project, open world multiplayer, with dynamic events and destructible, and roaming ai, environment spanning a wild range of machine? That's gonna stress any workflow, and technology and gameplay will be stress.
Stuff like hit feedback (don't play the particle until the server agree but play the animation + state rollback mecanism) are like complex problem that affect the architecture very deeply, good luck trying something similar in unity, even if you are a veteran!
They don't need to compete with their customer, they need to dog food.
I mean when they make the demo scenes, it results in a back and forth between demo and development teams, and changes are being made. Them making a game would also have similar benefits, I'm guessing?
The logic behind people asking them to make a game is that it often feels like the people that make Unity, do not use Unity all that much, and as such it's easy for them to overlook super annoying, badly designed, or broken things about Unity. I'm not convinced that them making a game is the solution to this, but hey, I can't deny it makes some sense at least on a surface level. At the very least it would be more productive than other things they put effort into, like the Icon Collective snaps.
Maybe the solution is something else, I don't know, it's their problem, they should solve it.
Lol! I bet you dont even touch these features. Unet photon or my own multiplayer dedicated server you name it. I am telling you, with UE4 Ndisplay you dont event need to code a game in multiplayer ways but just configuring the nodes right.
If you are wondering what I do, I work at a virtual production studio using UE4 in downtown los angeles. while previously developed mmorpg in shanghai using unity.
My grief came from years of developing with unity and realizing ue4 a couple years ago.
From shader forge, amplify editor, uber shader, write your own shader... how many techs become deprecated because unity is just reaching UE4's standard that has been set years ago?
I do hope ecs one day crushes UE4, and I expect that. But some unity users are like blender fanatics, engines are tools, my coworkers know unity and ue4 as much as I do, we never choose an engine and claim faction. and unity is a tool with many flaws compared with ue4.
My grief comes from inactivity and loss of focus of unity. I started with unity, and if unity treated game dev's well I'd be happy. But almost everyone in los angeles HAVE to transition to UE4 at some point in their career now it seems. If unity was that robust, we could have focused on one tool
I'm not sure what your emphasis on location is all about to be honest.
Also going to someone s feed classic act!
Maybe you can tell me how to implement a multiplayer game in unity?
Theres no solid choice! Photon is a is still vastly inferior to ue4s built in multiplayer that can be abstracted to blue printing and be a part of design!
In ue4 its smooth as butter and for virtual production we could just simply use ndisplay to do distributed rendering and vrpn for inpu for multiplayer. Thats why I am asking for cluster rendering. Even with that, VRPN Input and event is another beast! UE4 has it all..
I bet these aren't even relevant to you. because they aren't common gameplay or mobile game features. That's why if u r doing mobile games or not going for next gen arts, u dont realize how weak unity is
Again, I m posting all these drawbacks because I wish Unity the best because this is a tool I have years of experience and time with. But for bigger projects, UE4 dominates like Maya and Mobu. Its like Blender home studio users think they are the best, but in any top game hub, you'd be surprised how much tools have been in Mobu and Maya for professional mocap/animation/game rigging. much less with Qualisys or Vicon
This post does not go that much in depth, UE4 is a lot more matured than what this post suggests.
ECS is another beast, but essentailly what Unity lacks is a streamlined workflow.
Even with ECS, to reach to UE4 level, Unity at the very least needs to provide very streamlined solution for at least: 1. AI (AI Planner is not good enough GOAP cannot replace BT) 2. Networking(too embarrased to mention) 3.Visual Scripting is a must(season C++ programmer appreciates how much VS abstracts. It alleviates a team, stop thinking as you are one dev....) 4.Chaos Physics (No More Instantiate shattered mesh or rolling our own or DestroyIT. Learn how much superior Chaos is) 5.Game Framework like in UE4(Playerstart, default game controller, player pawn, even global variable & state management as an Engine Feature! This comes with UE4........ a hundre studios rolling their gamemanager.cs or using scriptable object is a nightmare. Quality varies a lot and none of them is UE4 level)
Jesus christ what studio do you work for that was using photon to make a AAA multiplayer game? Normally in a professional environment making a game of the scale your talking about, you make a multliplayer game the same way any studio makes one, you write your own networking.
Even commercial UE4 networked games like ark etc have publically stated that they have rolled their own netcode.
The problem doesnt sound so much as "unity cant do multiplayer games" as much as you guys as a team of professionals for some reason decided to rely on asset store solutions instead of writing your own networking yet expecting a AAA quality networked game from it.
Yes, DOTs networking isnt here and Unet was crap, but you can still write standard networking using sockets or similar techniques just as games and applications have for a very long time.
Also what you say about unreal and networking is rubbish, Im a user of the engine and its in no way a case of just wiring some nodes and bam you have a multiplayer game, anything made like that will be super unscalable and run like crap. Your making it sounds far more intuitive and straight forward than it is, anyone who uses unreal forums regularly will see that unreal networking is also a sore point for devs of that engine too.
While Unity is increasing their subscription rates, Unreal is giving more for free.
Basically, making games (Fortnite) provides them with a massive income stream, making this all possible.
Obviously, tt seems their strategy is working better than Unity's and I'm sure Unreal is far from done making strategic acquisitions.
Also, pretty significant regarding ArchViz (Unity is also targeting this market), the features of Unreal Studio are being rolled into Unreal Engine and will be made available to everyone for free.
Big warning signs for Unity here...
Not free, 5% of your revenue per quarter per platform. For hobbyist games, yes unreal is better value for money on that front. For anyone making a semi-serious commercial offering, 5% per quarter per platform from gross revenue is not "free" or even close to it.
No, Unreal has not changed the '5% per quarter per platform from gross revenue' to a higher percentage. So, the new content is offered at no extra cost while Unity has raised their subscription rates.
The audience for whom it is free are probably the most vocal about being impressed by free stuff, though. Its clever; the appeal of the constant stream of giveaways is primarily to the hobbyist tier, and 'dabblers' most of all.
Don't people realize that Epic try to destroy the whole industry. Start from Steam and now Unity. When the devil plan complete, it's too late to stop. Unless someone gonna save the industry once again.
We have to do something now!
That is a good point, also highlights why I am probably fighting a battle not worth fighting
Steam isn't good. You know that, right? You know that Valve having a monopoly is bad, right?
And see, there's the thing: every game has dramatically different needs and, by and large, those needs are being met. "Oh, any game!" seems to imply that this isn't the case. Should they make a mobile f2p title? An MMO? A triple-A scope third person shooter? A dialogue heavy RPG? Pong? LSD: Dream Emulator? A localized simulation of a galaxy cluster?
Which one of these games is bringing the practical benefits people are expecting that Unity isn't getting from directly working alongside studios like they already do?
At which point is the resource expenditure of starting up what amounts to an internal game studio going to see practical benefit over working with existing studios? When does that genuinely high cost, which we all know is high because we're all developers here, really start to pay off? Why is everyone operating off the assumption that Unity doesn't use their own product in any meaningful capacity during development?
See, it's nice to go "oh, Epic uses their own engine!" but it really seems to be a case of "the grass is always greener on the other side."
Yea bro go with godot or blender game engine or better yet roll it ur own.
Because that’s how many aspects of Unity feel. Features get started and abandoned and left in unusable states, new features reintroduce issues we’ve been over already (LWRP has gamma/linear issues still, welcome to 3 years ago), and editor speed is getting worse and worse.
Them not using Unity makes more sense than them actually using it and going “this is fine, let’s organize a Unite now”.
What made you assume we used photon for our multiplayer game? Let's not be a prick here and just list solid facts shall we
Photon and unet are from far far ages. For our mmo its complete custom backend and front end unity plugin using protobuf and later turnt into capandproto. The engineer of it lives here in los angeles.
If you are doing mmo, sure you need a custom backend at some point. But ue4 is good enough as a start until the studio finds success. Nobody starts developing a backend. We buy matured solution used for multiple projects or start with an awesome game until it has some traction
U r wrong about ue4 multiplayer. It is extremely robust. It's better than any solution on the market available for smaller studios or indies. And just telling you a lot of fortnite functionalities are in ue4 multiplayer.
The infrastructure is just solid. Can you build a fortnite with unity? Yes! Do it! Better yet probabaly go for godot and blender. It's all about that open source love for indies isnt it
What's giving you the impression that similar things haven't happened in UE4, the engine that has gone through so many lighting system changes in the last few years that they near rival the abandonment rate of every Unity feature that was abandoned?
Again, this is "the grass is greener."
Before the threadlock.
Unity employee: Sire they want us to make a game with Unity?!
Unity CEO: LOL what's next on the agenda?
Unity employee: The 'Dark theme' being released for free.
Unity CEO: Never! That was our greatest idea, increase the paywall!
The list of awesome ue4 features just go on and on and on. Artv1 and its standard skeleton, is already one step forward than unity. Those are not engine features but cultivated a very streamlined technical art pipeline.
This is a game engine that is dedicated to build awesome games. Not some ecs running simulation reflect thing. If that's ur thing, cool.
But unity is losing the battle as a game engine and further in the entertainment industry like film and animation
Nothing, I never said anything specific about Epic, I haven't seriously played with a recent Unreal version to have an opinion on what it is and isn't these days.
I'm just saying, people keep saying Unity should make a game because Unity feels like a piece of software that isn't used by its creators. Whether or not it's true it's another matter, but it's something that should be addressed.
My thing is making games and, shockingly, Unity's been doing that pretty damn well.
In near future, we have no choice but Epic. Atleast Valve done nothing much to their engine and improved their service. Let's the other companies do their job. The whole industry chain together in peace and harmony. Then Epic came and own everything. This is the end game.
Problem is unity has serious holes and gaps in its game. Having to use something like rewired is just a joke. Volumetric lighting, shader forge. Fundamental things like this not being a part of the engine prior to render pipeline is just one of the tales.
Now ECS will break everything. But it's okay. I just hope unity focuses on game, animation. And film. Otherwise I'll leave it for the archviz app developers
That's an incredibly subjective opinion..
I've used CryEngine and Lumberyard, both used in-house. They were genuinely the worst engines I've ever tried to work in. They were cohesive if you were trying to make one or two different genres of game.
UE4 is used in-house, and it's still not really a far sight better. It's also still far more difficult to extend than Unity, even with source access. Getting things that function outside the current scope of UE4 is actually a real pain in the ass. It's better only if you only need its out of the box features.
Unity's dominance isn't because of its tutorial availability or asset availability. Those things are both side effects of Unity genuinely being the best option outside of "just roll your own."
Just announced, Epic buys Unity! and drops all HDRP, URP, and almost everything else, the first thing they deprecated is ECS, BUT, makes it free for all mobile development, let me find the link... give me a minute
Also Unity now only runs on AMD, nvidia is not supported, wow!
Ah yes, how dare Epic not let companies do their job by...
Just lemme check my notes here...
Ah, here we are! How dare Epic not let companies do their job by offsetting a large chunk of development costs in exchange for temporary exclusivity.
Like, you know that's what's happening, right? Epic is offering minimum sales figures and sometimes outright funding for games that are released on EGS as temporary exclusives. Do you know what that does? It allows game dev companies to do their job going into the future because they have money.
We got Unity I you haven't realized.
None respecting self AAA studio will go with hands tidying third party assets, like networking solutions. Specially knowing how violently their are. If going AAA, expecting having serious budged and expertise on shop floor. Or just making walking simulator.
Yet Unity still allows us makes tons of games every year. Including 50% of mobile market. If not for fortnite lucky shoot, which wasn't bright at all at the beginning, wouldn't be so a loud crying from UE fanboys here.
So to me, looking into success of many Unity games, you really struggle with utilizing tools you have available on hand. Probably do to low programming expertise, which is why so much you are focusing on drag and drop features.
Serious AAA studios...?
That's done without mmo before, but I was just amazed how freaking easy this is with ue4.
Like you said. Everything ue4 can do, unity can. But ue4 its just intuitive and have u covered whatever logic design problem pops up. Like animation blueprint... ability system.. in unity how to talk to game manager is already a freaking insane topic.
Ue4 provides a freaking robust game framework upfront and a base class to extend. With unity I remember how we had to refactoring for days and test for days to introduce scriptableobject into the pipeline...
Agree with what you said I highly encourage ppl to at least try a few other engines. Not blender game engine or godot but something robust and decent. Then choose ur weapon of choice.
In fact unity is still great for not animation or ai design heavy projects. And way better in 2d that's for sure.
Do you know what? That leave me the question. Why I still use Unity today? I don't want to talk about that. I try both engines but.... Yeah...
Or it's because they went free first.
I don't know, I'm of the opinion that Unity, ever since 4.x, is slowly becoming a worse engine (not in all aspects, but still). I don't know about the competition, maybe it's even worse. What I'm seeing is, available stuff are less usable and reliable than they used to be, expanding is shakier than it used to be (since less used APIs break/change often) and editor experience is in free fall.
Ue4 dominates in even AAA studios. The problemb is that they are moving into film and animation. A lot of ups crave a streamlined process of making a game -> realtime movie -> rendered movie.
Ue4 connects gaming, previz/realtime art, all the way to render in something like maya arnold.
I love the old unity days. But it just saddens me they are not so focused on their game tech
Only if you're comparing UE4 and Unity. If you compare UE4 and in-house engines it's a whole other story.
Also, you're complaining they're not focused on tech when they've added the new render pipelines, a whole new VFX setup, a built-in shader editor, and a whole host of other things? Hell, you're even complaining about tech that's being added like DOTS.
The pain must be heal. Unity will rise one again. I try to skip this month. And see what coming next month. Thank you everyone and see ya.
Wait till RTX is fully supported then you'll eat your words!
That's true. I agree with most of ur points. But things like behavior tree, a better animation system, previz.
These really need to happen and not become an asset on the assets tore. If unity does not perfect itself, but only provides barebone features like ecs, I dont know I'd that's a good sign.
Ue4 is less performant than ecs, but how integrated it is is the killer point. The day when unity dev's stop building the wheels. And needs to provide the same way ue4 enables transition from game -> film
464709"]Only if you're comparing UE4 and Unity. If you compare UE4 and in-house engines it's a whole other story.
Also, you're complaining they're not focused on tech when they've added the new render pipelines, a whole new VFX setup, a built-in shader editor, and a whole host of other things? Hell, you're even complaining about tech that's being added like DOTS.[/QUOTE]
RTX will save us all!
*when there's finally a reasonable amount of end-user adoption
**which won't be for years
***because most people don't upgrade their graphics cards until they absolutely need to
****because they're wicked expensive
*****also it probably won't happen until well after console level raytracing adoption
******because RTX really isn't a draw outside of super tech oriented folks
*******because the end-user cost:benefit ratio really isn't there
********because the leap in visual quality is getting more and more difficult to justify after over a decade of chasing the photorealism unicorn
*********which makes RTX look like an incremental upgrade more than anything else
Unity I'm afraid is literally way behind in RTX support, here is the new 4.24 feature set for UE4
Ray Tracing Features Updates (Beta). We have a number of new improvements to highlight - some are:
New RTGI method that delivers faster results. (Experimental)
Enable it using r.RayTracing.GlobalIllumination.EnableFinalGather 1. It supports a single bounce and samples per pixel should be set to 8.
Improved instancing support for Instanced Static Meshes and Hierarchical Instanced Static Meshes with improved efficiency for large worlds.
Multi-view support for Virtual Reality devices and split-screen. (Experimental)
NOTE: denoiser support is not yet in Preview 1. It will come later in the preview releases.
World Position Offset support for Static Meshes (enabled with per-Actor setting)
Improved multi-bounce ray traced reflections with better support for area shadowing in reflections using samples per pixel greater than 1. We’ve also added support for SSR fallback when using the command r.RayTracing.Reflections.Hybrid.
Niagara VFX support for Ribbons.
Screen Space Global Illumination (SSGI) (Beta). We’ve added beta support for dynamic global illumination as a screen space effect. Currently, it can be enabled using r.SSGI.Quality. Use a value between 1-4 to choose a quality level. It’s intensity and tint color can be adjusted using the Post Process Volume > Rendering Features > Global Illumination category.
Material Layers (Beta). Will enable you to combine your Materials in a stack giving you similar functionality to Material Functions except that is supports the creation of child instances. Existing documentation on this feature can be referenced here.
New Atmosphere Fog Component. We have a new AtmosphereSky component which adds a physically based Earth-like atmosphere. It can be used to create exotic worlds and provides a ground view with and aerial perspectives, including ground to space views for planetary atmospheres.
Burley Subsurface Scattering. We’ve added the Burley algorithm to the SSS Profiles Asset. This method is more physically accurate and aims to improve the quality of skin shading and simplifying setup using physically based material properties. This SSS model targets high-end skin rendering with cleaner, more accurate falloff.
Enable it in the SSS Profile Asset.
The Editor Preview Level should be set to Cinematic
It requires Temporal Anti-Aliasing to be enabled.
For existing content using the standard SSS Profiles, it should require minimal changes to your existing content.
I am pretty happy with how hdrp looks now. The problem with unity again, is that it's not aggressive enough in film and animation like ue4. Many studios and hardware in la do not support unity. That makes ue4 the only option if a studio is doing a game that has movie plans.
Wait, you mean if I get an RTX card I'm stuck moving to unreal, never! I'll wait for unity to catch up.
RTX light probes are <3