What will Unity be like in 5 years time I wonder...

I've been working with Unity since Unity 5, before the render pipelines, before the new input system etc...I finished a project that took about a decade, and battled through upgrading to 2017...I think 2017 was the most difficult to upgrade to, I forget to be honest. One of the most welcome additions was that Unity does all the Input stuff for you, so you can hook up a Playstation or Xbox controller these days and it just works. That wasn't always the case.

I'm looking into taking on a new project that I anticipate will take 2-3 years, but may take longer...

A huge lesson for me with Unity, programming, software and hardware in general, is that when you're looking at taking on a project you know will take that much time or even more to develop, and you want to stay recent and up to date with the software you're using, you need to try to anticipate where everything is going, and we have to just make assumptions on what the future will look like, since it's a lot of time/money to be investing over a sizeable amount of time.

I'm basically fearful of one situation. Years from now, for some reason, I make a bigger project using the UWP and end up regretting my decision, as you can't switch from one pipeline to the other apparently easily?

Things like that worry me. If you were taking on a larger project and knew it would take years to develop, what would you look out for, how would you prepare for this?

 

In five years it'll take ten years to enter play mode.

 

Speaking of which, the part I haven't been fond of using Unity is the Library cache folder. If they could fix that thing not becoming like 30GB of disk space that's the biggest issue. The bloat is real at this point.

 

I've not had too many issues with that but I admit that's because my dev drive is 8tb.

 

Kind of a good idea these days.

 

Probably better off with a 2tb nvme

 

Given the amount of projects I'm usually involved with and the amount of platforms and builds I tend to have, I'm genuinely thinking of some sort of network based solution.

That was my old solution. I tend to have multiple projects that all need to be stored locally due to contract work and builds for each. 2tb was fine until a few years ago and now the only thing that would offer up enough space would be a RAID of multiple 2tb sticks running off pci-e, which is a much more expensive prospect.

A few of the projects I work with locally are over 250gb in size.

As to be on topic, my actual thought about what Unity will be like in 5 years is not a particularly glowing prediction. Fragmentation between in-engine features and packages will have continued to grow more severe, making onboarding new developers more difficult as they're forced to wrangle with what their starting projects will require; the SRPs will finally have full feature parity with built-in, but will still have fundamentally incompatible elements like how they handle post processing; more packages will be announced, implemented, and deprecated, but with each of them only containing some necessary functionality; documentation will continue to languish; Unity will keep trying to branch out into new industries, only to leave the core engine to suffer more and more due to poor management.

There's a distinct lack of focus in Unity's ongoing development and it's causing plenty of problems, honestly.

 

There is an 8TB NVMe drive.

https://www.amazon.com/SABRENT-Internal-Extreme-Performance-SB-RKT4P-8TB/dp/B09WZK8YMY/

 

Oh boy, the price is hefty too.

I suppose same cost of 4x 2TB. + saving slots.

 

@Murgilod In Unity3D's defense, Unreal Engine is 40GB just for the engine and is doing the same exact thing with trying to branch out and take over every industry.

By the way I'd be terrified of SSD failure with an 8GB drive.

 

If you're deving on LTS, then Unity in two to three years will be much like it is now. Everything takes them 5+ years now to get done and I don't think anything currently in the works will be particularly engine breaking. They're big on backwards compatibility above all else these days.

 

Maybe nano tech have changed how we work in unity by then.

 

I'm calling the police for your attempt at assault with that price tag.

 

You sound like you are already anticipating changes and planning appropriately. I don't think you can predict exactly what will change. As for things requiring more memory or faster components, etc. that is always the case and the technology generally keeps pace. Things won't suddenly take 5 times the RAM if RAM isn't 1/5th the cost.

I'd suggest extreme modularization so you could swap out some of the things you've built for whatever you decide to replace it with. That also should help if you decide to try an upgraded subsystem and it fails to meet your needs.

Also don't underestimate the importance of keeping copious notes. I don't know how many times I've decided to implement something again and forgotten exactly what settings to set. Many are so situational that you can run your app for days and not encounter the combination that fails. Oh... and add whatever you can in the way of validation-type stuff. I for instance mark my Inspector properties as [Required] (I use Odin) and if I miss something it won't build.

 

A 40gb editor install is a far sight different from the library folder of a project install ballooning to near that size and Unreal's efforts to expand have been much more measured than Unity's, who are also trying to branch off into a million different service options, usually by buying out companies and making their products worse in the meantime.

 

My predictions:

- Unitys main RP will be URP in 5 years and HDRP features will be ported to URP continously. You will be able to toggle them on/off depending on your target device.

- The long term goal for HDRP will be to be treated like an addon for URP.

- BiRP will be dead.

- Asset Store publishers will mainly focus on URP.

 

The roadmap is a good way to see what will be Unity in 5 years, in my opinion, it will be almost the same, good tools for small/mid scale projects for indies, great graphics (with some work) but behind unreal, i hope better documentation (like some years before) to be honest in 5 years i think unity will be in a delicate situation in terms of popularity, i hope i'm wrong

 

that is basically what is happening today

 

My prediction:
In five years some college kid will release an editor tool that is roughly the equivalent to Unreal Engine's Nanite and Lumin as it is now (too little too late). It will cost $350 and all of the big studios that use Unity will adopt it. Unity will buy them out and provide it for free to all users, but then development will stop completely. As soon as Unity 2027.2 comes out then this solution will no longer be compatible and will give you hundreds of error messages if you try to install it. People will say that you should leave your project in Unity 2026 LTS and never upgrade iif you're serious about finishing a project.

 

For a gen 4 nvme thats pretty reasonable.

 

Funding. I'd definitely look out for funding.

 

My general advice is use the best tooling you have now and not where the tooling will be in 5 years time.

5 years time is unpredictable in tech; for example look at how fast AI generators have progressed in just the last year! And I remember 5 years ago when every major tech company said that we would all be using game engines and creative tooling completely in VR by this point in time and 'flat screens' would be a thing of the past.

Admittedly, Unity does not have a great track record with keeping to roadmaps so I would not base your choice around what is being touted for 2024, 2025, 2026 etc. Look at current LTS, current Tech Stream and the very next Tech Stream release and evaluate the state of the tooling and how suitable it is for your project.

Also, when you shift from prototyping/pre-production to production you will want to lock to a version of engine and stay to it unless you absolutely have to upgrade for a bugfix to a hard blocker or a platform requirement. When I visited studios whilst working at Unity, the studios that version locked typically shipped whereas the studios that chased the latest versions multiplied the amount of issues they had (due to regressions, etc) and often massively extended their development length (which in-turn burns a lot of money)

 

My expectation for unity in 5 years:
- The editor will get even slower and more buggy.
- Out of all the QoL improvements people ask for the editor, unity will bring a few of them.
- More DOTS, Burst, ECS. Games in runtime will be even faster.
- A newer version of .NET, games in runtime will be even faster.
- But the tools unity provides (animation system, VFX system, baking lights, etc.) won't be usable outside of tiny games and demos.
- UGUI will be a way to go in many cases because the UI Toolkit lacks features.
- More pointless buyouts of other companies instead of creating tools.
- More services.
- I'm expecting financial problems in the future.
- This means cutting a lot of positions.
- This means tools that are now barely alive will get abandoned.
- The unity engine will keep losing to UE in terms of creating hyperrealistic graphics, especially for games.

 

Do not choose Unity if you plan a really long development cycle.

 

Is that 'in 5 years' a typo?

 

I'm only posting safe bets.

Trying to predict what will happen with RPs, input systems, or graph-based tools is just too difficult. It feels like everyone knows there are problems, but the developers, consumers, and management have different ideas of how to deal with those issues.

In the end, the roadmap lands somewhere in between. Who gets the biggest slice of the cake depends on corporate politics.

 

I think the implication is that you very accurately described the current situation, but I guess that is what you're saying, that nothing will change, things will remain the same or will continue towards the same downwards vector they currently are going.

 

Think the only thing REALLY worth thinking of on this long timescale is DOTS. So if you plan a game that will significantly profit from that, because it uses mass of enemies and stuff like that, then keep a close eye on DOTS and hope that they stick to the roadmap conceptwise. That's where most change is to be expected.

Most other things will just get gradually a bit better/prettier (URP/HDRP), possibly with one or two drawbacks but that's how it is. At the moment, otherwise the only other groundbreaking thing that is foreseeable for Unity might be Nano Tech (something similar to Nanite) but that's 3rd party: https://forum.unity.com/threads/nan...ndering-for-hdrp-urp-and-built-in-rp.1292223/

The switch to C# NET from Microsoft instead of Mono/Il2cpp should be completed roughly in 5 years, but that does not bring many new features, just performance - especially editor performance.
That said, the next gen PCs should have even less issues with the editor. Just don't use office notebooks designed for Word & Co.

For the Runtime it's expected in 1-2 years and the editor comes later, if I remember right from that large thread.
Albeit I am a bit confused why the runtime is less work than the editor since the runtime has to take into account so many more platforms than the editor.

 

MORE. HAIR TECH. DEMOS.

 

Now I have a feeling that 5 years from now DOTS will still be unfinished.

 

In 5 years there will be no Unity. The way things are going I see Unity completely collapsing.

 

xD lol, so much faith in the tool you are using

 

Am I missing an /s? Unity coasts on inertia of existing pipelines, knowledgebase and community size that'll take a lot longer than just 5 years to demolish even with the current unfocused Unity leadership. The core of the engine is basically frozen in time and is maintained relatively well, I don't think that'll change in 5 years no matter what happens in package land or what they acquire next.

The way I see the situation, it could go both ways. Either Unity pulls it together, cuts off old legacy fat and refocuses on replacement technologies where applicable or they continue to come up with ever new MVP poorly documented packages which further fragment the community to the point new devs will start to choose engines that don't require pHD in package manager dependencies and incompatibilities. And even if the latter option comes true, it'll be a very slow process. Changing engines is costly.

 

5 years.

 

Unity going bankrupt in 5 years is not completely out of the question. I think it would be easier to say that in a somewhat larger timeframe, because sometimes the process of bankruptcy itself takes many years, but still.

 

Well, if they went bankrupt, I wonder what would happen to all the licensing. Would the released source code be able to be used commercially at that point, if the company doesn't exist.

 

(for the record, I don't think they will go bankrupt in 5 years, I think it's more possible that they will continue pivoting away from video games, until they find something new that works, and they will continue downsizing).

But if it does happen, we would all probably be F***ed is my guess.

 

The tool is too significant to just disappear and be given up. They would be bought up by some other company who would automatically uphold the licences. Of course that company could establish a different business model etc. from that point forward and cancel support for certain versions if they wanted.

We shall see. Please do not call the 5% layoffs "downsizing" tho. Unity hasn't really shown downsizing yet.

 

Then we will all be version locked like so many here advocate

 

We would be version locked to whatever is the latest version we can pirate, because I doubt a bankrupt company would do anything to keep licensing servers running. Remember a year (or two?) ago, that for a day no-one could use Unity because some server along their data mining / licensing chain was acting up?

 

I read an article not long ago about the Elder Scrolls VI and how Bethesda is anticipating the game will be relevant for a decade just like Skyrim was. I think it goes to show how much things have changed. A couple of decades ago 10 year old games were considered extremely antiquated but now it's just the way things are because development cycles for massive AAA games are insane.

I'm guessing the Unity will continue to develop with their current pace. It would take a disturbance of some sort in the game industry to change that. Those are hard to predict so I'm not going to try, but I think because of the long development cycles game industry has an economic interest in trying to keep things stable.

 

I don't know how that's relevant for Unity because hardly any big AAA productions use Unity, and Unity is terrible for long development cycles.

 

I think its more likely it will be acquired by one of the biggest tech companies than go under entirely. The share price is heading downwards constantly and most of the execs are continuously selling their shares (all this info is public and easily google-able). Also looking at competition in UE5 dominating majority of markets Unity wants to move into and Godot coming for the smaller projects with a much better tech foundation. I think its fair to take these into consideration when trying to predict the next 5 years of the company and product.

There is alot of value still in Unity; it just needs to know its strengths and focus on that. Cut the random products & features that the community won't use or need and focus on improving why people gathered to it in the first place and what they use it for.

 

Or how bethesda have changed.

Their attitude, however, could be reasonable and related to NVIdia's CEO attitude who claimed that Moore's law is dead. During 1990s, 2000s and up to 2010 gaming industry was riding on semiconductor improvements, where each couple of years you could expect a new generation of tech more amazing than previous one.

If t he CEO is right, then this is now over, and we've hit the ceiling. Or, at least, a plateu. So, basically, imagine a world where current technology level in GPUs could remain unchanged for 30 years.

Anyway.... time will tell how it turns out.

 

Alot of gamer centric analyses here. Unity is the go to for industrial interactive 3D applications across multiple hardware configurations. Unreal will not fit on embedded devices nor work with RTOSes [Real Time Operating Systems]. The integration of Unity with various hardware platforms, industrial installations, automobile, military equipment and simulators continues apace in the background. It's .NET scripting language ensures integration with current business systems. Unity is doing just fine and has a great reputation in high end big buck industries.

 

Well, to be fair, Unreal is reaaally close to photorealism. With enough effort as a Demo scene it's feasible. And even Unity did it.
So what could come next that demands more hardware? 8K? I have big trouble believing that'll become the standard in the next 15 years. Heck, 4K is still pretty much enthusiasm-level despite existing for over a decade.
Maybe VR will be the next driver, but with fov-accounting rendering, you don't need extreme power. Maybe one more doubling of an RTX 4090 for photorealism.

So I think the progress will be rather in the tooling. Making those photorealistic scenes that currently require enormous effort needs to become feasible for indies.
Nanite is supposedly a step in that direction due to it simplifying some of the usual LOD-related workflow and reducing worries regarding polycount.

Honestly I'm happy if the big studios are forced to focus on intricate gameplay again instead of visual fidelity.

 

I think the main reason regarding Unitys most recent (mediocre) stock-market performance is due to the global economy situation + shortsellers ruining the market. The best you can do against those shortsellers is to hold for as long as necessary.

 

Unity's stock is underperforming the market or the average tech company, so not sure if that holds water.

Also, it doesn't matter what the reason is, their stock is down a ton, they messed up their ads revenue (which is most of their revenue), they were forced to merge with an ads company to stay afloat, their "core" audience (if we assume that's game developers, which it probably isn't) is more dissatisfied than ever and the biggest part of their userbase (kids and small indie devs who will never release a game) is one wind shift away from moving to another engine.

 

Living world.

For example: a life-sized city with 12 million unique NPC each with their own agency, unique voice, looks, clothes and work schedule.

Once you achieve life-city, the next frontier will be life-sized Earth.

Then making it multiplayer and fully destructible.

 

Well, that is a matter of content though, not so much hardware resources (on client side) because there are only so many NPCs you can see and interact with at the same time.
Maybe that will be the field for "creative" AI ala ChatGPT, Dall-E, Stable Diffusion etc. to shine as they'll help creating such NPCs and other content.
One day we shall have a game like in "Free Guy"!

That's a point. Together with realistic graphics (aka not voxels) that is currently still unfeasible due to hardware limitations (not software/content) and probably will stay like that for a while.

EDIT: Maybe let's add realtime proper water physics to the list where you can take a glass of water, fill it with water from the ocean and then pour it nice looking and without obviously behaving differently than in the ocean, onto a table or the ground, splashing and everything.
Though we are pushing more into matrix teritory than fun games here