UE5

You can always use FMOD or Wwise, one free for one release per year for indies, one free for a limited number of assets.

 

Best not mention networking then. Oops.

I don't really want to see new tech. I'd love existing tech and tooling to be optimised more and improved over and over. Funnily enough that's exactly what happened through the UE4 life cycle.

 

Much as it'd be nice to see tooling and tech optimized, recent events like the transition from the asset store to the package manager being a total mess have left me less than enthused about Unity's promises going forward.

 

The choice between Unity and Unreal is not a complex question at all, it's completely black and white. Add C# scripting to Unreal Engine, support the language as a first class citizen, and I'll jump ship and abandon Unity forever. C++ and Blueprints alone are not good scripting languages for my needs. I'd also like Blueprints to be improved as well.

 

While this isn't official it is being actively developed.

https://github.com/nxrighthere/UnrealCLR

 

It doesn't seem like/as far as we know*, C# will ever be an official, first-class scripting solution in Unreal 5, but Unreal Verse, or whatever they end up calling it, looks quite promising.

This post is 5 months old. If you click the Twitter link, you can see screenshots of the scripting language in action. I'm certainly intrigued here.

https://www.reddit.com/r/unrealengine/comments/kfne3e/first_look_at_unreal_engines_new_scripting/

 

They're taking their time. I think that's fine.



Compared to that Unity appears to be suffering from ADD.

 

At a glance it has similarities to Pascal.

 

I looked at screenshots.
I thought it meant to be C#.
Or I misunderstand something?

 

Its early examples of UnrealVerse.

 

It is actually more similar to python.

Because this kind of thing is a typical python layout.

Code (csharp):

1.  
2. if a:
3.     blah
4. else:
5.     bleh
6.  

Significant indent, ":" to denote sections, and no line separators.

Pascal had "begin/end", obligatory dot at the end of program, ";" for line separators and marked sectiosn for variables, interface and implementation.

 

The topics are everywhere on this thread. lol.

I don't often jump into debate threads, but I want to add my 2 cents.
I might be the only one here, but I enjoy Unity and Unreal Engine. Both engines have pros and cons. However, It's always been about the journey and destination for me.

Let's say both engines are Automobile GPS Navigation Systems, and you are the driver of the car.

One has Voices Included, and the other doesn't but has Street Names.
One has the Faster Route, and the other has Heavy Traffic Avoidance.
One has zoom-in and zoom-out views, and the other can list nearby gas stations and hotels.

With all of that said, their goals are the same, to guide you to your destination. The journey is the fun part. The destination is the reward. The route you take is entirely up to you, the driver.

I'm making a game that I may never finish, but I am enjoying the journey along the way and meeting great developers as I progress.

I'm enjoying Unity's new features from beta and alpha versions.
I'm enjoying Unreal 5 early access features.
Is it a sin to like both?

Cheers,
-crawling back under my rock of development.

 

Yeah, the only way to enjoy deving with Unity is if you don’t intend to finish.

 

its all about having fun but it is important to me to finish the games in a timeframe because i want to make some money and make more than one game too.

Now of course the type of game is biggest factor in engine choice. What Ive learned in the past week is that the months of problem solving i spent making one of my projects in unity would have been a week of not problem solving in unreal, adn a lot of money i spent buying assets to help solve those problems wouldnt have been a cost in unreal because the tools are in built.

So yeah, its not like unity is a human and i have apersonal vendetta against them, but when it comes to finishing my products the engine actually matters a lot. Like one is a go and one is a no go.

 

I was trying to be optimistic. The goal is to finish (and earn money) but sadly for me neither engine does what I want to achieve out of the box and has to be developed regardless. I'm having fun, so hopefully, fingers crossed, I will finish.

 

It's skooskumscript with slight syntax changes. I used skookum fairly heavily and can see the same keywords and concepts.

Also meant to add that skookum basically tries to copy ruby features. Such as range operator .. ? after bools and a bunch of other stuff.

 

Looking at nanite... do you can't make the latest horizon zero dawn with it?
- heavy foliage,
- transparent objects,
- thin geometry aggregate close together, like grass
- dynamic mesh like sand terrain that deforms

.... that's pretty limiting...

However i expect photogrammetry workflow to stary running with this, scan and drop and all around.

Apparently the main contribution is an efficient patch based workload, for decimation relative to pixels, for rigid geometry, with streaming friendly format.

 

Right. I think they're working on all of the above, thankfully. Early access and all that.

This last one though:

dynamic mesh like sand terrain that deforms

I haven't played with it yet, so I wasn't aware of this. Can't we still import any terrain mesh into the engine like we saw in the demos? Or did you mean it deforms while the game is running?

 

Wished Unity had something like Nanite. That to me is even more important than their Lumen system.
But before Unity even attempted it - it would be nice to have all the stuff such as HDRP and Terrain system actually working right (Grass, Trees), making the SRP's run good, bring in DX12 up-to-date (workable with better render times or equal to DX11), DX12 always been slower for me.

Kind of curious if it's possible for asset developers to make tools like Nanite. Not sure why not, seems like a doable task.

The only thing I'm not understanding with UE5 is how they are handling Batches. I never hear anything about batches, only geometry (Nanite) and Lumen.

 

Not really, u use default rendering for those and enjoy nanite performance boost on static meshes.

 

I finally tried UE5 and I'm baffled how every problem I have with Unity just doesn't seem to exist. Especially regarding the UI/UX:

• The startup time was like 3 seconds. Meanwhile I'm waiting for 2 minutes in Unity to create an empty project.

• The Editor feels so smooth, responsive and is really nice to look at. This is such an amazing modern design that doesn't only feel extremely well to control, but it's also really organized. Going back to Unity feels like going from Windows 10 to Windows XP. The design of UE5 is lightyears ahead of Unity in every possible way. The way the tabs open, the way windows are slightly animated, the material previews, how it feels to drag objects into the level... it's hard to describe, but it just feels so much better here. I honestly feel like Unity's UI and entire UX needs another complete overhaul from the ground up. How can everything in UE5 feel so smooth while Unity is so clunky?

• Blueprints and the material editor feel amazing. They feel so organized and, again, the controls are smooth as butter. Bolt feels like a children's toy in comparison. THIS is how a graph editor should feel like.

• Everything just feels modern and like it's exactly where it's supposed to be. Even the preferences are fun to control, since everything's so smooth and organized. And this is not only true for the basic functionality, but every single feature I've tried so far. There are no half-baked packages, deprecated features or unclear options. It feels like a complete package that gives you everything you need right out of the box. There's no Render Pipeline confusion, there are not dozens of ways to do the same thing, and I feel like every feature here is ready for production. UE5 seems to give me a very clear way of doing things that just works.

I swear to god, if Unity wasn't better for 2D, I'd be jumping ship right now. If we had an engine with UE5's user interface, usability and design philosophy in a 2D engine, I'd never touch Unity again.

 

Valid points. But I hope unity team is not going to waste manpower for improving UI/UX editor at this point.

 

I agree, I think unity's UI for the most part is pretty acceptable, if you're using dark theme.
I'd even say I like it. UE4 UI was pretty bad, so this upgrade was needed and a welcome change for the engine. Not so much for unity imo.

 

When talking about UI - at least when I am - it's not so much about colors and shapes, more about logic and ergonomics.

A good UI guides you to where you need to go even if you aren't 100% sure. It actually can help teach noobs, and help people who know what they are doing move quicker with less stress along common routes.

A good UI never makes you search with your eyes or reach with the mouse. When you click A, all the common contextual actions are ready and waiting within easy reach.

In total, you don't notice a good UI at all. You are 100% focused on your work. You can work as fast as you can think, and at the end of the day your eyes and wrist don't hurt.

Also, in the same way we as game designers manage player satisfaction by delivering steady micro-victories, a good UI accomplishes the same thing by giving us the layout we need to work fast and efficient. In short, a good UI is a pleasure to use on a moment to moment basis.

This is important. To me, it's more important than tech upgrades, especially for the class of developers who will never make use of tech upgrades until the point they are common and out of date.

Personally I find UE4's UI to be very nice. When I open editors for the first time - like the animation editor for instance - I find it easy to find my way around and setting up my state machines seems easy, logical, and fast.

It's been awhile since i had my fist experience opening unitys animation graph. I only touch it on rare occasions but every time I do I have to go and look up online what to do. It just doesn't make sense to me.

Maybe I am a big dummy, but it took me 20 minutes to figure out how to pay for more storage through collab so I think maybe whoever is designing things at unity is just on a totally different brainwave than me. In the end it was such a hassle it gave me time to seriously consider if I really want to pay for me storage from collab. I went to GitHub. There again, the github desktop app is so easy to use I was able to accomplish my task in no time and continue on with my work. I didn't have to look anything up, the UI guided me exactly where I needed to go.

Some people here might look down their nose and say, "well this stuff isn't made for morons so who cares if it gave you trouble?" But the thing is, there is a lot more morons than otherwise, so if you want to make money, you design stuff for morons.

 

Well just a few comments from a blundering new comer, with a handful of weeks under their belts. I apologize if the following is somehow mis-represented.

Unity's Interface, if that's the right word, is terribly confusing.

1. You are presented with various boxes at the beginning, none of which is clear what is the most sensible option, to confuse matters, some versions say for example 'VR' appears to require a special branch, how is one to know this?

2. The package manager, is all over the place. Again, the more clicks you expect the user to execute the more likely they are to throw in the towel. I understand it's there to provide a leaner experience so it doesn't ship with bloat, but UE5 has everything there where you need it. No or hardly any external dependencies required.

3. Unclear, conflicting information with user input, animation etc, we haven't even touched the graphics pipeline yet, but we can assume it MUST reside in a similar state to the rest of unity, we've read horror stories with their lighting pipeline, to an extent where they seemed to have sold and then re-bought a light mapping solution that was plagued with errors from day 1?

4. Scripting (comments from my daughter), needs to be re-engineered from the ground up. UE5 appear to have gone for their own 'domain specific language' in unreal verse. It appears that c# and dots, which doesn't even look like c# and thus defeats the point has been horribly hacked together to afford functionality. On smaller games, this might not be an issue but as you scale it could indeed pose a problem.

5. Unity's 2D tooling seems better than UE5 at the moment, although if UE5 wished too they could pretty much just knock the ball out the park with that one should they invest the time and resources.

6. UE5 pipeline, from everything through to asset creation to the end goal, is tightly integrated into their engine, it's a seamless process, with unity it seems like a constant battle.

Overall, unity has avoided a 'KISS' approach and have gone for a cobbled, keep adding approach and it is failing for anyone seriously considering a 3D game. It's very uncertain what path they should take next, starting from scratch is never easy, but a mis-match of promises and tooling is making life more difficult than it could be.

 

Does anybody here remember when all that we wanted was the ability to use shadows? I miss the simplicity of those unity days.

I know I'm an old fartage now, but I do miss things being simple.

 

primitive != simple

 

In my opinion, what Unity needs to do is get away from the yearly releases and instead go back to the old way of doing things, which is how Unreal still currently does it. Do a major engine version every couple of years where you can overhaul everything and then do regular smaller updates in between those major versions.
Back in 2014, Epic upgraded from UE3 to UE4 and it was a complete overhaul. They've done a lot of updates in between, and now they're moving on to the next big version which is Unreal 5. It has big exciting new features, the editor is faster and has a new UI, all that stuff. Unity used to do it that way with the major X.0 versions (Unity 4, Unity 5, etc.).

It's done this way for a reason. As you develop the software, sometimes you have to do spit and glue patches and fixes to make stuff work in the short term, but aren't necessarily the best solution in the long run. Having a big new release every so often is important because that's when you can do spring cleaning, fix up the internal architecture, refactor old stuff, and so on. Start fresh. It's also the point where you can establish a solid plan and vision for the engine for the next few years from a bigger picture standpoint.
But when you're in this perpetual cycle of smaller updates and no major deadlines, then it's harder to do that. It's harder to focus on long term solutions because the next update is always around the corner and it has to have something. It keeps you stuck looking at the smaller picture.

The old release cycle also makes it easier to collect customer feedback since all the attention is on this one particular version and the focus is on making sure that one version is a complete upgrade in every way. Right now, feedback is sort of all over the place--when someone complains about feature X, which version of Unity was it with which rendering pipeline, with which version of package manager feature Y? And the fact that people (like myself) are forced to stick with older versions of Unity because of stupid stuff like randomly removing the Display Resolution Dialog window (without offering a replacement) is ridiculous.

And not only is the old way better from a software standpoint, it's also much better from a marketing standpoint because then the new version release can be treated as a major event. It gets FOMO working in your favor, the whole watercooler talk thing. It's similar to the marketing difference between a TV show that gets new episodes every week vs a Netflix show that gives you all the episodes once--in the binge model, the buzz dies down shortly after release, whereas the weekly model keeps the buzz going much longer.

Well lets be honest, nobody really cares much when a new Unity version comes out. Everyone knows there will be a Unity 2021, 2022, 2023, etc. and roughly when it will come out. It's not a major event, there's no fear of missing out, just business as usual. Unreal 5 however, IS a major event because the last major version like this came out like 7 years ago and it's bringing in revolutionary new features.

And the thing is, Unity desperately needs this kind of overhaul. Unity has a ton of half-complete features that have been in development for literally years at this point, and there doesn't seem to be any particular rush to actually finish them. And the mess is starting to pile up.

So how about this...make a new overhauled version of Unity. Treat the yearly releases we've been having since Unity 5.6 as Unity 6, and do a clean break and call this new version Unity 7. Get all the existing features production ready, streamline the UI so everything makes sense, make everything stable and performant, refactor stuff as needed, implement larger scale changes based on customer feedback, etc.
Then a year later, release Unity 7.1, then 7.2, etc. Move away from regular updates with fixed deadlines and instead release updates when they're ready and have features worth the upgrade.

 

Unity is not for the bedroom coder any more. It's now a behemoth that caters only for large teams, with all the expensive tools, and the ability to draw on said teams' expertise to accomplish something.

I don't think Unity itself knows what it has lost.

UE5, well, it's UE5. It is something that Unity wants to be. At the moment, Unity can do it all, except offer a simple way to do it.

 

Except Unity doesn't cater to large teams because most of the systems are too simplified.

Making a larger title with the collection of plugins that weren't designed to work together? It's far from efficient. The more complex game becomes, it's more and more crucial to have an engine in which dozens of components work well with each other.

Unity is currently the only major licensed engine where we can't use blogs/Reddit/discord to read and discuss how the engine works because source code isn't publicly available. There's no accumulated public knowledge on how the engine works.

That shows the problem of Unity's identity. It was an amazing engine for small teams and games. But in last years they're trying to pursue every area of gamedev (except perhaps AAA open worlds), AEC, cinema... without any coherent vision.
It's sad that Unity didn't simply decide to make money on their own games. Any financial success with the game would allow avoiding this gigantic pressure on "making money on everything possible, as fast as possible".

 

Totally agree. It's useless just making everyone excited for the next release, like 2021, and then have barely any features. Just have major releases every couple of years and that will be the best.

 

Here's a meme I saw on Unity's subreddit, which sums up all my thoughts about Unity:

https://www.reddit.com/r/Unity3D/comments/ef8jrh/unity_in_2016_vs_2019/

Edit: Sorry for the broken image, just attached the link

 

I doubt money is their problem, they make a ton from their ads solution, built into unity.
Making that type of cash with games is not easy.
They have a lot more employees than epic, for example.

 

This doesn't seem to be the case anymore. Online sources report Epic having 10k employees and unity having 3k.

 

I highly doubt that number of employees being low is Unity's problem.

 

I don't really agree with the UE love-fest. The frame rate of the editor is terrible, and you still have to slide the UI across to read field names, etc. Lumen uses a lot of screen-space data, so indoor scenes will randomly go pitch black, and it ghosts, constantly. Nanite has no settings, so you will have to be ok with 20+ million polys on the screen. It's preset for next-gen and the algorithm is too complex to let you tweak any cascading (as mentioned by the developer).

Unity just needs to focus in on things that games will need, otherwise I might aswell just make my next game in Roblox.

• LoD generation
• Pathfinding
  • Patch up NavMeshComponents and integrate it.
• Modern in-game UI
  • Unity's new UI is just HTML/CSS but missing most of the features you want.
    
  • Supports soft masks/anti-aliasing
  • Shaders that work in-world
  • How did Genshin Impact make theirs in Unity?
• World creation
  • Procedural terrain generation
    • See Houdini/Map Magic/etc
  • Procedural object placement
  • Auto-handled nearby terrain object enabling/disabling based on tiles
  • Procedural sky/water/weather
  • Tesselation
  • More fog/particle options that don't require HDRP
  • UE5's outdoor demo would be less impressive if you could just generate it in 5 seconds inside Unity
• Visual Scripting
  • Revive Bolt 2...

And I would still like them to add VFX/Shader graph to built-in...

 

On double checking, are you sure the 10k employee count isn't for epic.com? Epic.com has 10k employees, you might've gotten them confused, as I did.

From what I can find, Epic games has ~5k employees.

 

Roblox is an excellent choice for someone wanting to get into game development for the first time, but it's a bit of a bizarre platform in my opinion because it's very lightweight to the point that it won't even let you use shaders but at the same time it's very hit or miss when it comes to integrated graphics.

Unreal Engine starts up with the second highest graphical setting regardless of whether you asked for maximum quality or scalable quality in the project window. The setting shown in the screenshot below only applies to the editor so if you're experiencing performance problems feel free to dial it down.

Spoiler: Settings window

I wasn't planning on commenting on the remainder of your post but I just couldn't help it after seeing this statement which is very much incorrect. You're only going to see 20+ million triangles if you're working with a scene that has the potential for that many.

In the screenshot below I imported a stylized environmental asset pack known as "Advanced Village Pack" and told the editor to enable Nanite on every non-foliage static mesh. The number of triangles rendered by Nanite was a bit less than 60,000 while the rest of the scene was around 560,000 (maximum was the correct camera angle here).

Spoiler: Advanced Village Pack

[EDIT: Before you look at the second screenshot I want to point out I know the camera angle is a bit off from the first one. Correcting it resulted in the above 560,000 number becoming 490,000. Nanite thus shaves off 210,000 triangles not 140,000.]

With Nanite disabled the number went up significantly to around 700,000 (ignore max here it was from angle caused by a foliage heavy angle looking at the sun and back). Staring directly at the sun resulted in a triangle count of zero since there were no static meshes.

Spoiler: Disabled Nanite

Static meshes that have Nanite clearly have an advantage. Without Nanite on them they were contributing 140,000 triangles more than when it was enabled. Between that and the way it optimizes other aspects like drawcalls (only one drawcall for a material regardless of the number of objects using it) I see no reason to disable it.

 

Yep. My mistake.

 

Ignoring UE5 for the moment.

I am more concerned that Unity has been relatively quiet since the end of 2019. We have had a handful of blog post that mentions their focus, several features that Unity and everyone else were excited for have been marked as deprecated with the mention that news will be forthcoming but there is no followup, DOTS development (entities part) is being moved either to 2020 or to 2022 but no real concrete decision.

The only reliable information I can gleam from Unity is by reading the github of all places... the roadmap is seldomly updated and the documentation has turned into a copy past for each version (recently there has been some life on the documentation that is fixing some of it)

Now the latest blog post mentions there will be no unite until later 2022.

From a business standpoint that makes me nervous because Unity in the past has always been very clear to me at least about their direction with the Unity technology base.

Granted Unity does not have to tell anyone anything at all. But it sure does help me plan and plan for my clients, and to help the asset store authors.

 

Could you please list them? Because this is news to me. I don't remember any substantial deprecation in recent versions. There are some on-going development on hold to finish other features, but no deprecation AFAIK.

What is that complicated in 'use 2020 until further notice'? What decision do you want other than that in a very experimental project? You shouldn't base any business decisions on ECS until it is marked verified. But you're free to play with it, obviously, if you want.

There was GDC Showcase earlier, Unity had a fairly detailed keynote. There will be GDC in this summer, I'm pretty sure they will have their keynote there too. They opened the https://resources.unity.com/unity-engine-roadmap where you can even suggest direction or vote for one not just read what they are planning. They are also frequently showing early access tools/development on the forums.

I don't see that they would be silent at all. But I guess we have different expectations.

 

Dots audio
Kinematica
Unity's global illumination
lwrp
Enlighten being removed then coming back to be removed again in 2024
AI navmesh system
Several more listed as being put into maintenance

From the DOTS thread

You must stay on Unity 2020 LTS and on Entities 0.17 for now. Future releases of Entities will not be compatible with Unity 2021 until the end of the year at the earliest. Upgrading to 2021.1 and using current or future Entities packages will not work and is not expected to work.

As soon as we can, we will provide further information on how we will continue to develop and release Entities

That specifically says only entities 17 will work on 2020.LTS, Future entities may work on 2021.LTS or they may only work on 2022.LTS no decision has been made.

The roadmap is rarely updated. It had features list as being planned when they were already released.

As for the GDC 2021 early this year not much was mentioned other than a continuation of the things mentioned in GDC 2020 with the focus on SRP, Netcode and visual scripting. Unity in the past had much more in depth presentations for GDC and Unite

 

This is what I am talking about here.

Nope. Not deprecated.
https://forum.unity.com/threads/dots-audio-discussion.651982/page-6#post-6875429

Nope. Not deprected.
https://forum.unity.com/threads/uni...haracter-animation.878761/page-4#post-6929078

Maybe, there is no deprecation. Although I give you that there is no clear word on this one. Although the wording of the Enlighten reactivation strongly suggest they didn't give up on it.

LOL. I guess you will mention this until the year 3254. It is by no means 'recent'.
Also the Unity Script and the legacy particle systems are missing from this list. Those were deprecated as well.
With that said, yes, this is a real very wrong step from Unity. (Not the deprecation, that's actually good, the initial plan was utterly wrong, the URP-HDRP lightyears better)

It is not recent either. And I don't know what to tell you if you're worried about 2024.

These are the "several ones".

Yepp, use 2020 until further notice. You shouldn't build business on top of an experimental package if this causes you substantial harm. If it does not cause harm, why would you even mention it?

 

When a feature is marked as postponed/suspend/etc until future notice its dead in the water for any developer to plan around it. Until development is restarted that feature is deprecated for me.

Exactly and this includes features marked postponed/suspend/etc

 

Ah I see. It is probably language barrier then. English is my second language and I learned a completely different meaning behind deprecated. This makes sense, sorry then.

 

No worries! We are all just wanting to make the best possible games/apps in Unity. Don't get me wrong I like Unity, The ability to create complex logic and test fast is awesome. I like the direction of the srps in 2021.2. Currently I wish I knew what Unity was thinking for future DOTS, Currently part of DOTS is ready for production 2020 and 2021 (burst, jobs, il2cpp math collections) these systems are great to use.

As for my use of deprecated you are correct it does have a specific meaning in the software world. I just use it more to the extreme so I and my clients do not get caught in mess. This is nothing specific to Unity I treat all my software recommendations the same. Anyone remember RenderWare and EA buying them...

 

They bought source code access and spend 100mln$. I'd say it's completely different engine at this point. This is also true of many AAA games made in Unreal but at least projects made by Epic contribute back to the source code.

Unity is losing ~80mln$ every year.

 

This.

I was working today and wanted to add animation to my UI in 2021.1.10f1 and somehow I picked the "Window>Sequencing>Timeline" option. I'm hit with a "To begin a new timeline with Canvas, create a Director component and a Timeline asset" > Create button. What? I have successfully moved from "Animation" to "Animator" but now I have to learn yet another way to animate?

I finally found the old-ish "Window>Animation>Animation" to get what I wanted, but that is friggin' confusing! How many different ways do we need to animate things? Unity needs to just pick one and make us migrate. Having all these options for everything – rendering, UI, animation, particles is ridiculous. Just pick something already and make it our only option!

I don't mean to be harsh, but several of these areas are 2-3 years in preview stage. That's not great for confidence that it will be finished before it's deprecated. With such a huge learning curve for these things it's actually easier to just learn a new engine that has an idea of where they are going!

 

Yeah, adding my 2 cents, 2D developers are kinda trapped in unity for now, except simple games made from static images which can be made in Ren'Py for example.

UE has some interesting (and expensive) third party tools in its store that are supposed to replace abandoned Paper2D, but lots of specialized 2D software doesn't support exporting to these tools, it usually supports exporting to Paper2D only when it comes to UE.

Godot is promising, but still lacks many features that UE and Unity have for years. I'm not convinced it can catch up either, considering the holy grail of Godot 4 is about as ready as Unity's MLAPI.

But even then, many 2D features remain experimental in Unity for many years. It's like Unity doesn't care if developers want to use it or not.

 

I like Unity. So now with that being said...I get the feeling a lot of the users are waiting for "something" to jump ship because the Unity process of developing has turned from a fun thing, struggling to figure out what your game needs and how its going to work, to how am I going to make this work in Unity. Honestly if I really sit and think about it, I use Unity because of some really high quality Assets that the community made and is selling on the Asset store. These assets if they were not there or if they were elsewhere (Specific assets I use for my games), I would not be using Unity right now I guess. I'm also on the verge to check out UE5 properly - what's holding me back is learning a complete new way.....but it's also an exciting thought. It's a sad state as Unity back in the day was the engine to beat.

People make the comment that Epic has so much money compared to Unity. There is a reason for that .. they are doing things right so the money follows. One has to wonder that if they decide to also do 2D if Unity will be used as much for "AAA" or big Indie studios even...and then where will the money come from. Then also even though I think Godot is still far from being real competition for Unity - it's coming, in the same way that 10 years ago people were dismissing Blender. People who want to make 2D games are going to Godot, Defold and or even GameMaker and Construct 3 - depending on what games they want to make. People who want to make 3D games are looking wide eyed at Unreal Engine now, and from looking at YouTube a lot of people are going over.
The one really impressive thing Unity has for devs is that it runs on virtually anything. So compared to Unreal and Godot they have that in the bag, but I think maybe tat is also what could be limiting Unity in terms of development update speed. They need to make sure all the features work and can run everywhere etc.

For unreal it's not even just the graphical fidelity etc. It's all of these amazing tools combined. The terrain and vegetation system alone is amazing, then one gets things like MetaHuman, Quixel, particles, destruction system, fantastic built in Multiplayer etc with it...there are a lot of out-of-the-box features stuff right there.
In other words it's difficult to ignore if each of those features saves you a bunch of time and makes it possible for me as an indie to do it and make it look good as well.