Search Unity

  1. Unity support for visionOS is now available. Learn more in our blog post.
    Dismiss Notice

Unity Roadmap | Unite 2023

Discussion in 'General Discussion' started by PanthenEye, Nov 28, 2023.

  1. PanthenEye

    PanthenEye

    Joined:
    Oct 14, 2013
    Posts:
    2,005
     
    Last edited: Dec 8, 2023
    DungDajHjep and mahdi_jeddi like this.
  2. TheOtherMonarch

    TheOtherMonarch

    Joined:
    Jul 28, 2012
    Posts:
    856
    -Speedtree 10 sounds Interesting. Hopefully, more version 10 assets will be available right now 9 and 9.5 hardly has any assets; most assets are still v8 which is the last time Speedtree did a major asset refresh.

    -VFX graph improvements are nice.
     
    Last edited: Nov 28, 2023
  3. IllTemperedTunas

    IllTemperedTunas

    Joined:
    Aug 31, 2012
    Posts:
    763
    It cannot be overstate how thoroughly unexciting any of that was.

    Trying to be less negative on these forums, but they're not making it easy. One would think that after all the years and billions in investment there would be SOMETHING to be excited for.

    I wish more people who have attempted to make games had more control over the sorts of things that are being worked on at Unity.

    This felt like a collection of cloud based monetizations dreamt up by an untested marketing team, which is frustrating because Unity has unlimited access to thousands of developers intimate with their platform who could give them endless ideas that would be fruitful to countless users.

    The levels of disconnect here are mind blowing, and it's kind of just sad at this point. Tired of lackluster features hidden behind technical mumbo jumbo to try to jazz it up. Tired of a new array of features that will endlessly suck development out of what could be worthwhile investments. Tired of this noise getting in the way of Unity getting on track and the endless bleeding.

    The real kick in the nuts was finally seeing some native WETA stuff and it just being some over the top muscle sims.

    I expected nothing and I was still disappointed. Guess tolerable iteration times are something to look forward to.
     
    Last edited: Nov 28, 2023
  4. PanthenEye

    PanthenEye

    Joined:
    Oct 14, 2013
    Posts:
    2,005
    The key is to not have any expectations.
    There is for some niches like VR and high end productions that benefit from rendering and VFX graph improvements and also those who do multiplatform development at scale. URP is also getting some of the HDRP goodies finally.

    The most exciting stuff is the last minute - the sneak peek of Unity 2024 or is it Unity 7 now? Less progress bars and CoreCLR port, although they mentioned migrating to .NET7, when .NET8 is already out.
    The muscle stuff is all Ziva but rebranded to Unity Weta Tools. The only true Weta thing is the Maya plugin called Wig for movie quality hair grooming.
     
    IllTemperedTunas likes this.
  5. Murgilod

    Murgilod

    Joined:
    Nov 12, 2013
    Posts:
    10,010
    .NET 8 has been out for two weeks.
     
  6. PanthenEye

    PanthenEye

    Joined:
    Oct 14, 2013
    Posts:
    2,005
    It's not like CoreCLR port is landing tomorrow.
     
  7. I'm pretty sure, if it's feasible and doesn't hinder the rollout in any major way, they will change the target to .Net 8. Otherwise you will have to compromise with .Net 7 and then they will work on the changes needed to make to bring up to 8.
    I'm not sure why is this a problem right now...
     
  8. Glader

    Glader

    Joined:
    Aug 19, 2013
    Posts:
    454
    Speedtree sounded like a great thing like 10 years ago. Has anyone ever gotten to actually use it for anything since its integration to Unity back then?? I swear, it's been absent.
     
  9. jiraphatK

    jiraphatK

    Joined:
    Sep 29, 2018
    Posts:
    276
    Around 59:00 onward,
    seems like we are getting asset database V3?
     
  10. 00christian00

    00christian00

    Joined:
    Jul 22, 2012
    Posts:
    1,035
    What happened to the animation system that was shown a couple of years ago that allowed seamless transition between complex animations?
    I can't remember the name.
     
  11. I'm guessing you're talking about Kinematica?
    It's currently in deep sleep and probably accidentally gets killed at some point because it's since 2020. But they say they are planning (at some unspecified point) to do something with it. They say it's dependent on the new animation system (probably DOTS-based animation library which should replace Mecanim), but they didn't even mention it neither in the Keynote nor in the Roadmap (they spent time on AI bullshit instead), best case, it's 2-3 years out for a first release.

    https://forum.unity.com/threads/uni...haracter-animation.878761/page-4#post-8859375

    My opinion: just forget it and use Motion Matching from the asset store.
     
    Last edited by a moderator: Nov 29, 2023
    Shizola and 00christian00 like this.
  12. 00christian00

    00christian00

    Joined:
    Jul 22, 2012
    Posts:
    1,035
    Yes that. It the only interesting feature I saw in a while. Thanks, I'll have a look at that for future projects.
     
  13. sacb0y

    sacb0y

    Joined:
    May 9, 2016
    Posts:
    854
    They're rewriting the whole animation system and thus aren't talking about kinematica nor animation rigging.
     
  14. sacb0y

    sacb0y

    Joined:
    May 9, 2016
    Posts:
    854
    I reccomend peeps watch the live stream they did yesterday, there's a lot to like in the new updates. It's not fancy tech but it is very useful stuff. And they answer a lot of questions.
     
    IllTemperedTunas likes this.
  15. IllTemperedTunas

    IllTemperedTunas

    Joined:
    Aug 31, 2012
    Posts:
    763
    Watched a good bit of this, encouraging. Yes you see burnout, and you see a small team juggling a lot. You can tell they're tired of working on the SAME ISSUES over and over.

    They're tired of having to explain "Yes, it would be nice if we unified everything, but the entire point of the render pipelines is to get different visual styles utilizing hardware in completely different ways and the tradeoff of that is that the two systems aren't going to seamlessly work like a switch."

    BUT, in terms of engine implementation and standards of workflow that have been developed we're all running around like chickens with our heads cut off. There needs to be more planning and design sense for inspector setups to alleviate burnout from these guys by working on how things "live in the engine", the pipelines for how these settings are added, stored and how they're interfaced with both in inspectors and in code.

    If there was more investment into best quality of life rollouts and setting up contextual menus that these guys could simply unload onto rather than them having to juggle implementation and functionality (sort of the antithesis of what these guys do on the graphics team) I think we'd have a much better tools and a team that wasn't so burned out.

    These guys want to keep everything free flowing so they can iterate and push the envelope and keep moving in new directions. and making cool new toys not endlessly hear how their work isn't good enough or even usable.

    Us users just want to be able to click somewhere in the darned engine and be able to turn on a quality ambient occlusion solution where we would expect it to be.

    The biggest bottleneck for Unity right now is not performance related, it's not quality related, it's the organization and how features are implemented, and until this problem is sorted it's going to be a constant time sink, and source of pain for users and developers alike.

    So often we think because the importance of key post processings, or runtime lightings, or whatever that code and equations are going to be the lions share of the work and burnout, but if you're constantly fighting against your team, the managers and the users for where to even house this stuff and you're perpetually trying to figure out "ok how does the user enable this, where does it go, let alone how do we make it easy to find?" time and time and time again, you're going to end up burning out on the most bare bones part of the pipeline and burning out before you got a chance to polish the functionality. Next thing you know you've got teams of highly specialized, capable people all burned out and not doing what they're best at.

    Just the sense i've gotten from all this. I could go on about Odin inspector and the clunkiness of built in Unity stuff. But the main point was, it's not cutting edge tech that's holding unity back most of all, they need to figure out their internal pipelines to make adding that stuff less painful ideally by having a team composed of designers and coders who were dedicated to working with teams such as this to come up with not just better means of implementing these system but just design sensibilities as well. This would enable us to also better implement quality inspectors and our own tools.

    I know it's been said endlessly, but if you want to see a real-time graphics pipeline done right, check out Unreal, "It just works". Now try to get ambient occlusion working in Unity, it's madness.

    Until the ideal workflows are found, and methods of putting this stuff together in Unity, every graphical asset is going to be another unique puzzle trying to figure out how to get it to fit into the hodge podge of other graphical features that are all created with the sensibility that they're a la carte. It shouldn't be a la carte, there should be features you take and leave and others core to each pipeline with care taken to how they're implemented, and care taken for the implementation of the implementation.

    It feels like before more work is done to these pipelines Unity needs to sit down and have a discussion, "how the heck do we add this crap in a way that isn't insane" How do we stop burning out getting feedback from our Users and spending most our time fighting with implementation?

    It's a snowballing effect, the people adding things to the engine and making usability and figuring out how to put things in the engine in ergonomic ways get better and better at it. The people you hired to make lights look like lights while being performant get better at doing those things and rather than putting their work out there and getting heat for tertiary design stuff, they get the praise for the awesome stuff they made they rightly deserve. A positive feedback loop where everyone gets better at getting better.
     
    Last edited: Dec 1, 2023
    Andy-Touch likes this.
  16. zulo3d

    zulo3d

    Joined:
    Feb 18, 2023
    Posts:
    756
    Agreed. I remember testing a newer version of Unity after a long break and having to spend an hour trying to figure out how to add a post process effect. I tried going through one of Unity's tutorials and I got the impression that the guy that made the tutorial also didn't know how to do it.

    "To enable ambient occlusion you must do this thing and then this thing and these things and then... Oh is that the time?"

    yeah whatever..
     
  17. sacb0y

    sacb0y

    Joined:
    May 9, 2016
    Posts:
    854
    On one level I agree Unreal is easier in that regard, but on a different level I like how much control you have. Thats a lot of why I chose unity in the first place, control. As someone who releases on a PC and android the setup makes sense to me and is very useful. But if you only care about higher end platforms the setup can seem strange.

    Infact the main problem I see from the wide range of assets I test and use is confusion on why these things are here. Like Beautify instead of using render features for it's stuff instead tries to do stripping on it's own. Maybe there's a benefit to doing things this way but to me it's counter to how it should really be setup.

    Maybe not enough people are focused on URP and such to really "Get" that sort of thing. But hopefully more streams like the above will remedy that. Also I think with the render graph stuff, in the future more things can be default out the box with less issues.
     
  18. tatoforever

    tatoforever

    Joined:
    Apr 16, 2009
    Posts:
    4,361
    @sacb0y,
    That freedom of choice leads to unscalable projects, because you are not forcing developers to follow any kind of rules.
    Same goes with programming, if you don't enforce practices and methodologies, you will endup with a giant pile of messy code hard to work with. I've worked with many professional studios using Unity. And oh boi is so messy and everything is all over the place. No exception, everyone working with Unity has the same messy pattern in their projects because of this same "freedom" of choice.

    The same could be said about Unreal engine but it has better guardrails and enforcing rules (eg the gameplay framework, engine settings, etc). In fact, it works very similar to Unity but in a more structured/organized way.
    As someone currently working on a custom shading model with Unreal sources, i can tell you all their shading models have optimized/approximated paths for all platforms and you have a centralized menu to scale your graphics settings, no need to create X profile here, or go to some obscure asset there which is the case now with Unity render pipelines. Heck, even your GPU instancing number for all your instances can be scaled in that menu!

    With Unity, every single system seems like it was designed by different people with different methodologies and usability for a different end-user. It requires you to create an asset that you can put "who cares where" all over the place as with Unreal you have a centralized place to quickly scale and change your game settings at anytime which is saved to the engine/game ini files (also centralized in a single place). Bonus, a built-in console to change absolute all engine/game settings (that is saved to the same ini files). And I won't even mention all the amount of tools you have at your disposal which allows you to focus really more on your game than the tools.
    Long story short, with Unreal everything is centralized and clean. The level of control and scalability is better, you can go from absolute high end settings to a lightweight Forward renderer (cluster light or multi-light single-pass, pre-pass or not) with a single click, without having to change anything. As others pointed out "it just works". We can dream about what the future will look like but at the end of the day we want tools and features now to help us build rich interactive games faster.

    Unity has become highly complex and cluttered and is not an "author once and target all platforms" game engine anymore. Many fragmentations in very important areas: rendering, gameobjects vs ECS, multiple input systems, the list goes on. Unity now looks more of a sandbox multi-platform tool, than a game engine.

    I don't think it's sane to compare Unity to Unreal, one is years ahead in almost every department, the other is trying to running catch a train in motion...
     
    Last edited: Dec 1, 2023
    AcidArrow, nasos_333, pm007 and 5 others like this.
  19. sacb0y

    sacb0y

    Joined:
    May 9, 2016
    Posts:
    854
    I've seen some of that, but the amount of devs I know who like to make their own thing rather than work with the unity setup often causes these problems imo. It's not always the case, but the people i know who try to work more to understand and extend native unity features and functions tend to have a better time.

    People have the choice and ability to do these things so they do them rather than work with what's there. Often people assume a system doesn't work instead of learn to use it, like how many devs i know who just avoided things like addressables off the bat and do some custom solution (or an older solution) instead when really there's no reason to do that. Sure, Unreal has better guardrails but it also seems like people are more resistant to replacing or rebuilding anything in it, or they more willingly assume the tool is ideal enough.

    And while unreal is scalable I've still never really seen a high end game truely both visually and performance wise scale to a lower end device like a phone or a switch. I see it a lot with unity though. I've done it with my own experience.

    Agreed, I'd say unity's biggest mistake was having URP and HDRP be made by separate teams. At least now that issue has been remedied and the issues that caused being undone. I think the yearly release thing has been maybe to blame too.

    I hope that's a mistake they've learned to not make again.

    For my uses as a game developer this is still the case, it's probably not the case for asset developers for now.
     
    nasos_333 likes this.
  20. tatoforever

    tatoforever

    Joined:
    Apr 16, 2009
    Posts:
    4,361
    There's a bunch of high end PS5/XboxX Unreal titles working just fine on Nintendo Switch. Will they look the same on all platforms? Of course not, but the scalability is there and is very simple (few clicks away in the scalability menu). You don't have that in Unity.
    Also, I'm not aware of any high end game made with Unity. Yes, there's beautiful high profile indie games made with Unity, but none of them have high end graphic features. Unity at this moment do not scale well, this is not a secret. Either you pickup URP which lacks many graphics features and implement them yourself or HDRP, which does not scale down.
    Again, back to my statement, you have to write the tools and features which means less time working on your actual game, more time spent on production, and so on.

    Not only for asset developers, for anyone using Unity in 2023. They are all affected by this fragmentation and painful never-ending transition. One of the many reasons why we are starting more Unreal projects at work and less Unity projects. We don't feel that old excitement about Unity anymore and it's sad.
     
  21. sacb0y

    sacb0y

    Joined:
    May 9, 2016
    Posts:
    854
    ehhhh, I've seen nothing good that didn't look like a potato, or have very rough performance. The fact switch users consistently feel shafted when UE4 games come to switch communicates that well enough imo. Along with how long it often takes to get those results.

    Other games might have some custom features and stuff but they're still unity games. It's not like the Unreal games couldn't do that. And while they're largely anime styled games from asian developers it's undeniable the quality they produce is impressive and I would consider them high-end for what they are and the quality.

    Like this which uses I think URP.



    I won't argue with that, but IMO there's not a lot I would say I'm missing on URP that doesn't have some solution like SSR. It's really only shadows that's the main problem with URP at this point. To me there's little preventing URP from looking as good as a UE4 game (not UE5 given some features mind you), but imo it likely would look better and perform better on a mobile device.

    I agree but I also think people reinvent the wheel more than they really need to.
     
  22. PanthenEye

    PanthenEye

    Joined:
    Oct 14, 2013
    Posts:
    2,005
    The edited video just dropped:
     
    tatoforever likes this.
  23. AcidArrow

    AcidArrow

    Joined:
    May 20, 2010
    Posts:
    11,482
    They don’t “like” to make their own thing, they are just tired of hitting walls and having to do workarounds for almost every Unity feature the minute you do anything that isn’t non trivial.

    Every time I tried to embrace a Unity feature, we ended up making our own replacement.

    Also:
    You like how much “control” Unity gives you, but also you should really try to meet it where it is with its restrictive half baked features? Where is the control in that?
     
    IOU_RAY likes this.
  24. sacb0y

    sacb0y

    Joined:
    May 9, 2016
    Posts:
    854
    I'm just saying from personal experience a lot of programmers do it cause they can not cause they tried to actually learn the existing functionality. I can't count how many times someone is describing something they're trying to do and I'm like "Oh that already exists", and the response usually isn't "Yeah I know".

    Even when it comes to the unity store assets I use, I generally only use assets that integrate well with unity rather than go their own way. Those generally present the least problems.

    Yes, control breeds complexity, and limitations breed creativity. I'll use the base thing, make some functionality to work with it. Usually goes fine. Less problems upgrading unity versions too.

    Now this doesn't apply to everything, there's some things that are aweful. Like cloth, i haven't touched it for years but for my needs it was aweful and i've heard nothing change. But Magica is glorious.
     
  25. AcidArrow

    AcidArrow

    Joined:
    May 20, 2010
    Posts:
    11,482
    What kind of features?

    We’re probably talking about different things, because I don’t know what programmers you’re working with that don’t know UGUI, timeline, lightmapper etc features already exist. Features that I’ve fully regretted spending any time at all on and making bug reports for, when instead I should have went full on in on 1st or 3rd party alternatives, rather than naively hoping Unity will make them not suck at some point.
     
    Starburst999 likes this.