Unreal Engine 5 = Game Changer

I didn't change my mind, because I never had my mind set on using specific engine in the first place.

I'm not loyal to either engine, and will use whatever works.
The project you linked was initially born from having to transfer scene from unity to unreal and then from desire to look how lighting will look in the other engine, provided the assets are the same.

 

That's not what I meant.

When you march distance field - IF you raymarch distance field, it is possible to have a situation where the ray runs in parallel to some mostly flat surface. Because raymarcher can have maximum number of samples it is allowed to do, this sort of situation drains maximum number of steps immediately. Because distance to nearest surface does not change, and because distance field reading determines how far the ray can go, the raymarcher ends up wasting steps.

Spoiler

Look at the point where the ray runs parallel to the box. Normally in this case it will be walking in small steps, because distance field is not directional.

 

The silence is pretty deafening from Unity.

Now would be the time to at least make a public statement. "Hang in there a little longer, we have something really cool to show you. We've had some bumps in the road, but this will be worth it, trust us."

I wouldn't normally say this about a screeching user base having an online tantrum, but I feel as though we are entitled to an update about where the engine stands as we have invested our time and money. Our futures are inherently tied to the Unity platform as develors and a great deal of gamedev is long term planning. What systems and pipelines should we be planning for the future? How should we structure our projects in anticipation of new tech? Should we bother?

If anyone has some cool videos of upcoming tech from Unity that would be welcome. But I feel as though we are having to temper everything with a looming sense of "Oh god, DOTS is going to be a pain in the @$$". ECS, DOTs, Visual scripting... all in the future, all intangible, all potential disasters. People using the Unreal engine are throwing a pizza party and we're over here fasting with the promise of kale in the future, "Trust us it's going to be good for you in the long run!"

I just don't get it. The old Unity was capable of producing fantastic, fun games and had little issue running on current hardware. Performance was rarely a bottleneck. Hardware is evolving all the time anyway, that's the great thing about tech, if you just don't demand more horsepower, your engine will always run more efficiently with new hardware. Who greenlit the idea of risking everything for gains that weren't necessary? Who decided the entire pipeline of Unity was a gamble on some unproven tech by people who never created a game before?

Could they ONE TIME do something that inspires confidence in the past 4 years? Amplify Shader was great, but that was Amplify creations, not Unity. The prefab update was pretty good, but not quite as good as it could have been. That's the one thing I can point at in many years and say "pretty decent job Unity".

If you take a step back, yes, there are some things that are working pretty ok. Unity hub to manage versions, the package manager has lots of potential. But all of these things show an inherent lack of vision. What is your engine unity? Everything you have done has been a cop out to avoid having an actual identity. There is an expression in gamedev "when you target everyone with your game, you target no one." So what Unity has done essentially is instead of work to isolate and destroy issues from existence by phasing them out of releases, is they have propagated them everywhere. Now you have a huge laundry list of things that are half finished in the package manager, you have a slew of buggy builds to choose from from the HUB. Ugh. So much hard work and functionality to realize an ill fated dream. And one has to wonder if they are chasing waterfalls elsewhere.

The state of Unity is vital info. Info that makes or breaks people's lives. I hope Unity is in good hands, but if the core of their future updates are riddled with usability issues, then a great many developers putting their faith into Unity are going to be at a disadvantage in the coming years.

I've rambled plenty elsewhere in this thread so i'll try to cut down on these diatribes. At this point I'm kinda just beating a dead horse and none of this is anything they obviously don't know internally. Hopefully this backlash has helped give credence to the more sane voices within Unity or in some way help put a fire under their ass.

The current state of affairs is no one's fault. We all want gamedev to be better for everyone. Here's hoping Unity has some cool stuff cooking that will prove us naysayers wrong.

 

For me, workflow > billions of triangles.

So far Unity has presented itself as the engine most apt at helping solo devs and indies get the job done. But like others I''m worried that UT is losing that "democratising" focus, chasing UE with a bunch of preview packages instead of increasing editor stability and fixing glaring issues with the UX. To their credit, they've recently told us that our ceaseless complaints have been heard and that these areas will see more love. Now UT needs to find a way to clone Aras. And they also need a few hires from the asset store dev community to add new blood / perspectives. But sadly, it's Unreal which has recently been contacting and poaching some of the best UT asset store devs.

 

I think Unity has enough talent onboard. It's about managing and executing a vision.

 

So do you want them to lie ? I don't know what people expect Unity to say more than a tweet saying "cool stuff from UE", and that wouldn't even be good marketing.
People expect Unity to say "psyke we were kidding these past 4 years, here's the new unity" ? We know the route they were/are taking, and it seemed amazing when they first showed it to us.

No, and that's part of why Unity made DOTS, have you played Syberia 3 or Yooka Laylee (just to point a couple of high profile games) ? They ran terrible. Changing pipelines was a must, the built in batcher doesn't batch anything +1000 triangles, having a bunch of lights in forward killed performance, etc.
Monobehaviors are amazing to throw a bunch of scripts and have game logic blitzkrieg fast, but throw a bunch of them and the project starts to crawl.
The package manager was a great idea to make the editor lighter and don't feature bloat (which is ironic how it now created package bloat).

I don't blame users for feeling anxious about the state of Unity, it's extremely understandable (how many years ago they promised background asset import ?

https://twitter.com/reduzio/status/1262021308335915009
Godot working trying to replicate UE5

 

Just an update about where things stand would be helpful, I don't think anyone wants to be lied to, quite the contrary. A lot of the frustration is stemming from tech demos that don't properly represent their tech.

People want to know the timeframe for these investments paying off, how long this transition period will last. What the emerging plan is amidst all the changes. This is in many ways our future, we have a vested interest in it.

You're mentioning a lot of GPU based slowdowns related to forward rendering and real time batching. I was unaware the DOTS migration were impacting the platform so broadly. Truth be told I don't have much knowledge of the back end for these systems.

So you're saying ECS and DOTS will have a huge impact on the performance of Unity as a whole since the backend is all C#? If so that's pretty great, I just assumed all the hubbub with DOTS was that if you want to go through a lot of headache you could experience performance gain but only for specific systems you set up to run with DOTS.

 

DOTS is having and will have a huge impact, because it is delivered to you whether you like it or not. If you want to use it yourself, you will be able to do that (even right now, but especially when more systems will have DOTS-compatible public API), but more importantly they are moving (or will move) their systems in the background onto DOTS as well.
Great example the hierarchy, you probably don't know or don't notice or if you do, the user's overwhelming majority does not, but the object-hierarchy is running on DOTS already, even if you use GO/MB (this is why the best to keep as flat hierarchy in the scene as possible -> jobs spanned per root object, if you have one root object, it means the translation of your objects is running on a single thread, if you have 4 root objects, the translation of your hierarchy is happening on 4 threads - if you have available on your machine, of course).
This is one of the reasons I don't understand this new uproar against DOTS, it is benefiting literally everyone.

One more:

You have a lot of videos on Unite Now lately, you can watch them and will know where things stand.
I don't see Unity responding to Epic's advertisement, it's not the Superbowl.

 

I'm nervous for Unity's future, but not because of Unreal. They are two separate worlds as far as I'm concerned and plenty of room for both. The last thing I want is Unity to schiz out and try pivoting to match an Unreal tech demo. Noone has even seen this thing in action, just a flashy video, and everyone has lost their mind like they've never been marketed to before.

 

I see now, I really underestimated the impact DOTs would have moving forward as it impacts the entire framework of Unity. Maybe Unity can do a better job of explaining exactly what DOTs means for the engine as a whole and how this will impact everything globally in a positive way as performance = better graphics and a generally more enjoyable pipeline to work with, faster build times, etc.

The long term impact of this system makes sense now, and why it's gating various standardizations of the engine as its usability is iterated on.

Edit: I've spent a solid 15 minutes online trying to find a write up or speech that in any way mentions how DOTS is having such a broad impact on Unity. I really feel as though they could have done a better job explaining just how important DOTS is and us "noob" users would be more understanding of the delays and why progress is taking a bit longer as the major shift happens.

 

But they are telling this all the time. They are talking about moving systems to DOTS day in and out. Almost every presentation they made about the overview of DOTS and nearby things, they explained this or at least mentioned.

I wouldn't expect them going around the forum and respond every single misinformed or intentionally misleading post just to make sure everyone knows. If someone actually wants to know, they will go to the Youtube channel and watches the presentations they made.

 

I recall seeing "havoc powered dots", and I saw the mega city video which was really cool. But from my perspective they kept going over the process of coding dots, and it seemed to me that aside from a few new systems like Havok, nothing was really changing. I can't recall them ever saying "This is fundamentally changing the fabric of unity, it allows our interface to run faster, it improves build times, it makes batching more performant, it improves the absolute core of the engine and ensures we will be competitive well into the future. But in order to bring these changes to fruition takes time"

THAT I had no idea about. It appeared to me that it was dependent on the user to extract this bonus from DOTS and that a few new added animation, physics and runtime features would be built on it if you chose to use them.

 

I was just reading recently that Epic Games has updated the Shooter Game tech demo for UE 4.25. The Shooter Game is one of the earliest tech demos for UE4, which included online multiplayer functionality in a futuristic city.

Unity should do the same for many of the old tech demos. Give us the Blacksmith, Adam, or Book of the Dead, updated with Unity 2020 and all the new packages. Why are they releasing new tech demos without updating the old ones?

 

I was aware of Unity having nested objects would exponentially drag resources because it had to crawl between all the objects and update them, I was not aware that now that "performance tax" has been lifted. Even if I don't have dots platform or entities package in my project ?
That's really cool to know.
I guess another DOTS benefit all features will be to keep using mecanim/animator but change the background to dots, so all animation data is batch processed.

I doubt even Unity knows that, I doubt when they showed ECS 4 years ago they thought they would still be in preview in 2020. And I agree it's not good.

 

Let's resurrect it then! During an interview at Epic Games China one of their engineers told everyone that the PS5 demo ran just fine on his laptop with an RTX 2080 achieving 40+ FPS at 1440p. The engineer couldn't remember the exact model of the SSD he was using but did say that it doesn't need to be strictly specced.

https://www.pcgamer.com/unreal-engine-5-tech-demo-pc-performance/

 

If it wants to compete with UE on the graphics front, Unity should throw a lot of money at Inigo Quilez and sit back.

 

After 3 days of learning unreal I'm impressed how complete it is. I like how everything has separate editor/window that work as tab like in web browser.

Spoiler: For example

For example after selecting mesh in Unity, in Inspector we get:


In Unreal after double click:

On the other hand I like how in Unity there's simply .fbx and .meta file next to it. In Unreal these 2 files are combined to single .uasset making whole import workflow more complicated.

 

I generally prefer Unity's implementation you've pointed out as bad here because when I click on a mesh in the inspector, my general feeling is "I want to adjust the import settings of this mesh," but not "I want to assign materials to this mesh." When I want to mess around with the mesh's material stuff I do that in prefab space because it just good separation of concerns.

 

I completely agree with you!

 

That's EXACTLY what I was thinking about?
Let me rephrase to avoid misunderstanding, so there is something to point at:
- SDF marching have a tendency to slow down around parallel line due due to constant distance estimation.
- This is a problem for you because there is a finite number of sample you can done.

I pointed at that he used a cone based tracing to solve the issue
- first pass stop too early at edges due to size of the sphere sweep,
- second pass use that to refine edges by spanning rays at the first sweep end
From a screen casting point of view, the parallel lines problem happen to the projected edges of shapes, debug visualization show they are the most expensives marching.

Spoiler: Screen shot

PERFORMANCES HEATMAP: red is low (<8), green is medium (<32), blue is expensive(<128)

regular sdf tracing at full rez


Tracing with cone trace


Further edge fix:


well I reached the limit of image to uploads, check the video to see the result:

 

Last years before of the announcement of Unreal 5 the development of Unity has been very slow, for example Kinematica look like never will be stable and the new progressive GPU lightmapper to replace Enlighten Geomerics GI system have many problems after many time of work, then imagine how many years can take Unity to have new tools to compete with Unreals Nanite and Lumen.

 

Just a side note, HDRP only supports 8, yes 8 layers max in terrain splatmaps and no grass shader.

URP - Point lights cannot cast shadows, that is laughable , seriously?

 

8 layers max implies that they're tying terrain splatmaps to hardware texture slots, which is not the best idea.

Regarding point lights... well, this sounds awful, actually.

I mean, at least they could try to resurrect hard stencil shadows. Those do not require 6 renders per point light and work fine with pointlights.

 

Why is this? (Just curious)

 

ARe you asking why it implies or why is it no the best idea?

It is not the best idea, because it severely limits artist for no good reason. The toolchain should support thousands of textures at the same time and should handle splitting them into manageable chunks with X textures per terrain patch, where X is hardware dependent. Juggling splatmaps about should not be artists job.

It implies hardware reliance, because number is power of two and is awfully similar to Texture limits that existed at some point of time. For example, DirectX 9 had 8 texture stages total, no matter what you do. OpenGL allows hardware-defined number of textures being used simultaneously, but the minimum supported number is 16, which would leave you with exaclty 8 stages and normalmap + albedo for terrain splats.

 

I don't have any issue with Unity implementing DOTS behind the scenes. And if they can convert the normal component architecture to ECS automagically during build or whatever, without affecting stuff like debugging, that could be fine. More performance is always better. Although having two competing workflows doesn't sound like a good idea in terms of learning the engine.

But if the component workflow is getting thrown out for ECS, based on what I've seen that is a big mistake. Workflow is key, it's what you deal with every single day, and the ECS workflow is convoluted compared to components. In that demo I posted earlier, they used three different scripts to move a transform in a straight line. It's also hard to imagine how it would be compatible with the intention of customizing the behaviour of individual similar objects, which happens all the time in game dev as opposed to simulations.

I don't have anything against ECS - in fact I might have a use for it in the near future. But I don't think that the ECS workflow should be compulsory, especially for beginners and the majority (I imagine) of small indie games which have no use for it whatsoever.

 

Can you get mixed mode lights with specularatity and no real time shadows? I thinks specular lighting is more important than shadows most of the time. Baked lighting looks really flat because no specularity. Mixed mode looks better but then you also get real time shadows.

 

Because for some reason HD terrain support only two splat maps.

 

Lights that do not cast shadows are very oldschool regardless of fidelity of your assets.

You're looking at Quake 2 level visuals at worst, and something in ballpark of Silent HIll 2 at best. Silent Hill 2 had specular highlights. It was a PS2 game.

Doom 3 had shadowcasting point lights. It was released in 2004.
Blade of Darkness had realtime point lights with shadows. It was released in 2001.

It is a variation of a 19 years old tech. There's no excuse for a modern engine to be unable to render those.

The effect of dynamic shadow is very dramatic. If you don't have it, you'll be limited to static enviornments on overcast day, pretty much.

 

One good point in the post on garry's blog was rendering speed. Switching to ECS shouldn't affect rendering. If it is affected, there's some sort of architectural problem.

 

Well in VR you cant have too many realtime shadow casters because its too expensive, but we want specuilarity because it looks good. Plus real time shadows are pretty ugly, more so in Unity and especially from point lights.

This is how bad point lights look like, (though see all that sexy speculaitry, thats what we want)

So I would love a mode with baked shadows with speculiaty but no real time shadow.

 

My thoughts on the video:

To me it looks like the scene is very bright in shadows, and it is not about shadows being blurry. I particularly dislike corridor, as it is strongly lit up by default, and that kills the effect of light from the door being opened. The way I see it, If a light is dynamic it should flicker or move otherwise it is wasting GPU power for nothing.

In your situation I would investigate possibility of using completely custom shaders for the scene in order to get specularity you want.

Unfortunately, I cannot give a more detailed advice, as due to pandemic I failed to acquire proper VR headset. (not getting the damn thing any time soon, I guess).

 

And if you do use 8 textures you can't use undo, a 6 year old bug btw, that's my point of them making a game using the tools they make...

 

Given this, the package pain thread, the garry newman post making waves on reddit + twitter etc etc some sort of official statement on what is happening to combat the increasing user pain and lack of confidence within community that has grown since post-5.6 era unity (which we soon I am sure will call the dark ages) would be nice from Unity.

I know there was a half-admission of problems due to packages etc recently, but its gone quiet again and now is the time to show that they are listening to their user base.

Its been a long time since I heard so many long-time unity users (And long time users of this community in general) sounding so negative about the current state of affairs.

FYI if any unity staff are using this, I think in some senses remaining completely quiet is only helping the effect unreals marketing has had, not dampening it. Read what people are saying here, on twitter and on reddit and its constantly been getting worse and worse.

At some point something has to be done to win user confidence back and in general prove that the company is not just purely investor led with no real understanding of its core user base anymore (which 100% is where I stand now in my beliefs after being relatively reserved despite high user pain the last few months).

Having someone from staff jump onto a sticky package thread and engage users was a great start, but that initiative to engage users seems to have completely trickled to a halt very quickly which just makes it more annoying to be completely honest. If your going to engage us engage us, dont keep doing this cycle of listening for a month and remaining silent for another 11.

 

That would be a weak response and the reply from user base would be, "yeah, uh-huh sure."

The issue is say things and not follow through, so the only appropriate response is to be quiet and follow through. Deliver on some of the more important things and no words need to eb said. That's the professional thing to do.

 

If they were able to do that, this thread would be far less popular.

 

Assuming we aren't the vocal minority, how many blog post where actually about game developer recently? Who is the audience of these posts? Why they spend energy reaching to that audience if there wasn't any return?

 

Just to answer that last question, you’d give them a tag-like component. Then system would then filter based on your ‘tags’ and run whatever script is relevant to that tagged entity.

For instance, you have npcs that all have the same movement, you just want them to path differently based on whether they’re a brute or timid type. So you give some a “Brute Component Tag” and others a “Timid Component Tag”. Then in your system(s), you find your relevant tags and update their movement accordingly

 

When UE4 came out, it was completely panned by the community because it was buggy, missing features, completely changed many workflows from UDK, broke so many things from update to update, etc etc.

But how Epic got ahead of it was by getting on the forums and opening up a livestream where they talked with the community about upcoming features and what they were working on. It was a seemingly tumultuous time for them when they launched, but their response to the criticism was to visibly act on it.

Unity is in the same spot right now with the switch to ECS, SRP, etc but their response largely has been dead silence. And just as you say, their efforts to get on the forums were great initially, but all that has largely stopped (some still do though, the Input team/guy has been incredible in terms of communication of current and future plans/ideas/wants). And whats kinda sad is they’ll do efforts like this - https://portal.productboard.com/unity/1-unity-graphics/tabs/9-visual-effect-graph - but the chances are so high that this will go unmaintained and forgotten about that i don’t even want to get my hopes up.

I’m not sure what it’d take for Unity to be more open and consistent with its community, but this seems like the perfect time to get it together and attempt to dispel a lot of the complaints and help people understand where they’re attempting to go with all these changes.

 

Maybe Unity's current trouble maker is corporate culture.
I was lurking unreal's GitHub and basically they approach their solutions this way:

They hire you, developer, to create meaningful engine features. If you don't, they 'let you go' quick. So you HAVE to create something, even if they refuse merging said thing into the engine later;
But, here's the thing, once you have created something, Epic will slap your name on it and make you responsible. You are responsible for the maintenance of that new feature, it's your responsibility to keep it working despite engine moving forward with changes. You take personal responsibility for your code.
They apply the same philosophy to community contributions as well, doesn't matter if contributor works for Epic or not, you "own" the feature once released.

That's why you will see very old things like the UE4 Shooter Game demo project updated to every new engine version, for example.


Meanwhile, judging by this video:

I feel like in reality Unity staff won't take personal responsibility for anything they do.

 

They probably can't take personal responsibility due to the internal corporate structure. Also cross team feature is probably easy to move the blame around (like they did for the standard 3rd person controller). Because apparently they had to ask the team of animator, and the conclusion was not worth it due to "broad" use cases, even though a simple controller is also as narrow and don't cover any use efficiently anyway.

And that's one example, we have seen it with the shader team too, there is a deafness with hand wavy excuse that make seems to make sense but only reinforce dysfunctional statu quo. They aren't optimized around finding solution to problem, they seem optimized to over engineer silver bullets and pushing problem that don't fit it on the back, which mean they get a backlog of ever increasing problems and an ever increasing new feature list to implement. Since there is no mentality of solving problem, both issues start supporting each other into stagnation.

The unreal engine is an example of one idea solving many problem, ie new feature and old workflow problem at the same time. That's how you win.

Funny is that old unity use to have that bent, lightprobe interpolation and surface shader are stuff they were leading with, not following or catching up. A lot of the mobile optimization in shader they innovate a lot to get the most of little. Now they have state of art academic paper that don't make it into the engine, or catching up with what everybody else is doing.

 

I've mentioned potential structural issues in the Packages thread:

 

???

- Roadmap? -> actively responded to the criticism that Unity needs more stabilizing releases
- Unite now - meet the devs (basically from every part of Unity where devs are)
- Basically Q&A in every meeting/interactive sessions
- They will hold online meetups as well (just attended for the California/SF one, but I guess they will hold for other groups as well).

I don't know how this is dead silence and not communicating with users.
(And I understand if you don't like what they have to say, but I don't think they can be accused with silence)

 

I've worked at several dysfunctional tech companies and competency is the exception to the rule, especially in gaming and especially in the current climate where other considerations plague these work forces outside of simply getting the job done.

Unity is in the unique position of being able to sustain itself on past glory, even if current plans of attack are failing, and that's scary. Dysfunction follows where excellence is no longer required.

It's easy for gaslighting to start creeping into your pipelines as cliques start emerging and then EVERYTHING goes to sh@t. This is why the sustained lack of delivery from Unity is disturbing, because all their lovey-dovey tech talks and demos look to be band aides for the universal problem that plagues these tech companies.

Productive work or criticism can easily be dismissed along tribalistic lines and points of contention no longer form around production strategies, but along the cliques within the hierarchy. Some push for the interests of the product, some push for their own interests or the interests of their clique for mutual survival in a demanding industry that they can't compete in on a level playing field, or simply don't want to any more, they've found a gravy train.

A bloat begins to occur across the board and meritocracy becomes a distant memory. Instead of a team effort to produce product the internal structure of the company becomes a warzone as manchildren fight over their toys, their fat paychecks, for their cush positions, and for their stature and influence, or simply to keep a roof over their head. This is about survival, and people will do anything, you can't really blame them.

Gamedev is a very fragile thing, hopefully Unity hasn't been poisoned by this self serving infighting that is so rampant in this industry, but it wouldn't surprise me if they were, they are a prime candidate for this to happen to them. Isolated, fragmented internally, no clear leader with a vision for the technical pipelines, enamored with their "message" over their functionality, and most of all, simply not delivering for far too long.

One of the biggest give aways is how "hip" and "cool" they try to be. A gamedev craves not these things, it wreaks of tribalism, to appeal to our emotions, to want to seem likeable to us. That is poison in tech, that is when you become more worried about what people think of you, than of delivering a product. And in chasing this dream of being liked, of being hip, you lose sight of what it is you actually do. That's not to say marketing isn't important, selling an image isn't important. But you can't let it consume you. You can't let it influence your internals. You can't start drinking your own kool aide. Remember the Sega vs. Nintendo situatoin so long ago? Sega had to be cool and hip. They had blast processing, Nintendo was boring and stale. Which company survived in the long relying on solid tech over image?

Unity, please start empowering the boring, technical people who have slaved away their entire lives to be competent. Stop letting those who haven't put in the elbow grease and have relied on their charisma and backhanded tribalism to wrestle power from those who would seek to make your engine competitive. But here is the problem at these companies... what reason do they have not to ally with the people they like. Why does anyone in positions of power give a damn about us making our game, or the company competing with Unreal. A great many of these individuals simply do not give a damn about the engine. They are masters of feigning passion. "those other passionate people will handle it, i'm just a drop in the bucket." They've conned themselves.

Being on the outside looking in, it's easy to believe they expect the same things from their engines and teams as we do. But they are only human, they have their biases, their friends, their weakness to be conned just as any of us do. These are the variables that define whether you breed your talent into something meaningful, or if you're just not cut out for it in the long run. Most aren't. It takes catching lightning in a bottle for these intense technical ventures to work out.

Games are made on a fanciful dreams, but at the end of the day they run on tech. Enough whimsical tech demos and warm messages. More cold, hard functionality.

F@ck I can't help myself. I said i'd quite yapping, but this new conversation point hits me in my soul. I've been in the trenches, i've seen the bullshit, i've seen the utter lack of ability for companies that aren't top heavy with specialists to allow things to go to utter dog S*** and how easily that happens if allowed to fester for even a moment. I'm no top tier coder, but i've worked with some, most of them are door mats. They'll sit idly by as opportunists ruin everything they've worked for thinking many of them have their backs. It's sad.

For a long time i've felt so incredibly grateful with Unity. It has been like having an entire team of coders creating amazing functionality for me, as if the engine alone was akin to having a team of professionals helping you with top notch tools to help you create something amazing. Lately it feels more like you're walking on egg shells, as if you have that unreliable teammate that you can't trust, and their inability to execute reliably is now a potential hazard towards your future success. There is no greater fear when working on an expansive technical product than to worry that success is out of your hands, that you can invest years and years of hard work into something and some terrible outside variable outside of your control destroys what you worked so terribly hard on.

Unity as of late has started to feel like that potential outside variable.

 

People are mad that Unity doesn't have a response to a product announcement that won't even be available until 2021, which is weird and honestly ridiculous. I'm not sure what they want Unity to say about this that isn't covered in this actually fairly comprehensive video from a couple months ago:

 

"Premiered Mar 25, 2020" - couple? Barely two.

 

When a couple gets married, there's usually two of them!

 

I gotta say that in the end some of the information that surfaced in this thread is worrying. Unity side information.

 

We’ve made all these complaints before. Unity recently answered us in some degree of detail. Let’s at least give them a little time to show us it’s not just talk before piling on the criticism again?

What this thread shows is that users want constant engagement with UT staff. That’s fair enough and it’s something that can definitely be improved on. Honesty is always the best path when it comes to marketing: they should tell us about the flaws, the problems, the hold-ups; ask us for suggestions. But marketing people don’t get that. They think the answer to criticism is to ignore it or to put lipstick on a pig.

To put this into perspective, I think things are improving software-wise. Management-wise (ie. direction) seems like another matter though. My view is that the buck stops at the top, for that. And the “top” is very hard to move (unless they get a pay out or a court case interferes). No doubt many Unity staff feel this way as well, but it’s not their call to make.

As for 8 layers for HDRP terrains... we’re getting a whole new terrain system soon and I’m looking forward to that. Regarding layer count specifically, new system or not I’m going to keep using Microsplat, which not only does away with those limits but also adds a crazy amount of optimisations. It’s a shame UT don’t feel it’s important to at least keep their top asset devs in the loop as they too are passionate and act as user representatives in many ways - they stand up for the needs of real-users. Guess who Unreal have turned to for help on their terrain?

Regarding point light shadows in URP... the idea is “performance by default”. Point light shadows would likely not meet that criteria with a forward pipeline. Expect point light shadows very soon, when deferred lighting gets introduced. So once again this is a marketing / comms issue, for leaving us “in the dark”, for not realising Unity users are a passionate and rowdy bunch who want detail, not rosy blog posts about what’s coming several years down the track.

 

I think people are more mad that Unity won't have stuff they showed 4 years ago, meanwhile UE in 2021 will be fairly stable with state of the art tech.