Unity Technologies: please make a game.

OK So what problem areas should Unity focus on?

1. MIDI
2. WebGL microphone
3. Multiplayer
4. ??
5. ??
6. ??

 

I would not be so sure about this.

First thing, size of the project.
Many developers were reporting problem with editor stability related to project size.
Even on my small scene when I turn on the gizmos it starts to run slow.

Second, the size of the scene and related issues with lights, occlusion culling, lightmaps.

This might generate some rational thoughts, like built in prefab scoped lightmaping and local light probes.

And the networking.

 

I have been following this thread for some time, I think that many of the assumptions here are not so valid as "why unity does not create a game".

That is too vague and doesn't solve the problem, unity is a generic tool that can be used to create any game, and precisely for that reason it has many strengths and at the same time some pain points... The same way you can't recreate fortnite in unreal because epic changed the source code to do it... unity has access to unity's source code, and it can manipulate it in many ways to create a game. And this does not mean that unity would be improved if the game was created, only that for this game or style of game this modification would be useful.

So creating a game doesn't solve the issue.

I understand that in fact unity has some pain points, or flaws... They could listen more to the developers and improve these pain points, especially for small teams or solo developers. Rethink some existing solutions that were not well implemented or do not have a good workflow.

 

No. You're putting the cart before the horse entirely and also focusing on the things that would have the least meaningful impact on making a game in this context in the first place.

 

You sound like a broken record mentioning and asking for Midi in every comment you make...

 

OK What type of game should Unity make?

 

Pay attention please: they shouldn't make a game.

 

Yeah. I am making what is called a rhetorical point..ya know for emphasis...just in case the subtlety of wanting Unity core systems created instead of rehashing a rehash of a rehash with shiny objects is lost on Unity Tech. We don't need a Unity game. We need tools to make interactive 3D software across a multitude of industries. If you ain't a musician or never owned a DAW based studio or used a Midi controller to control animation in a 3D app it is perhaps beyond your mindset to fathom that request.And mic access on WebGL.. shameful their answers back in 2015 when it was requested and the reply was that is not at all on our priority list..then later in 2018 started and dropped because it didn't work everywhere all the time.. Want a Unity game? Play GooBall.

 

Heck.. If they got the first three working that would be just dandy. The MIDI integration beyond controller input would allow very large scores to be loaded in a very small file size. But let's prod Unity to use resources making something _some_ people will play instead of use. Makes no sense to me.

 

A handful? Last I checked their are over 100000 MIDI controllers on the market. I do not want to use the mic and controllers together. They are just two issues that popped up over the years. Unity was all ready to add MIDI and four people naysayed it loudly and they dropped it. They dismissed the utility of mic access on WebGL and years later on the same thread started it internally and then dropped it because it would not work in all browsers all the time. So..yeah..want mic access..port it to every platform and no sweat. Except in WebGL you have to jump through hoops, add dozens of defines, kill crashes after building for 45 minutes and it found another call to a mic. I am working on an intent engine and natural language processing brain for industrial and automotive 3D interfaces that monitors and controls systems. It may be niche but these products will be a part of millions of machines. I don't need gimmicks. I need solid bug free tools and proper cross platform accessibility.

 

Last I checked, this was a thread for gamedevs to voice their concerns with the engine in context of game development. Apparently, you don't have similar problems. Good for you, but that doesn't mean gamedev related issues are of less import than your niche use case.

Also, it is not just one or the other. People able to work on MIDI controller support are likely not the same ones working on current engine features and packages. We're discussing the current state of the engine, while you're pushing your feature requests Unity consider low priority or technically unfeasible at the current time.

Create a new thread with your feature requests and stop derailing the discussion.

 

This is a thread about FPS gamers wanting to divert Unity Tech resources to a pet project they think will be so awesome..like Unreal ya know! . I am against that. Yer cancel culture gambit failed pal. I am on topic.

 

That's not the case at all. You're just attempting to minimize game developer problems and to highlight your own, which have nothing to do with game development this thread is about. Perhaps you don't understand game developer problems because... you don't develop games.

 

You should probably let the team know to add that as the next button after the "Make MMO" button that I am sure is on the way

 

Are we reading the same thread?

Nobody wants Unity to do this because they think it will be awesome.

That want Unity to do this because they think it will *suck* and they want Unity as a company to see how much it sucks.

 

We joke about this but what if Unity worked on an automated/procedural world building system that could create a world and allow for plugin assets for a specific genre/style.

Isn't one of the biggest problem with game development creating enough content for gamers on a limited budget.

Even AAA studios are recognising this with people leaving older studios like DICE/EA to create studios dedicated to solving this problem of game development.

And this is before we get to the Meta verse where people will want to be able to create their own worlds with a press of a button. And to everyone fixated on the Make World button check out No Mans Sky and Minecraft.

 

To be more precise.
We want unity to create awesome game but we want them to see for themselves how much development sucks.

 

Concrete case:

Before i tackle anything, I read extensively manual, script reference, example of implementation, state of the art, spend some time over thinking the problem to cover all cases, THEN I choose the simplest way to implement things as I realized most of the time when you don't know, you tend to choose the most convoluted path, i don't like reinventing the wheel.

I'm working on refactoring a GI proof of concepts, it's simple, it is read some textures, blend and save the result.
So I made a first version that turned out spaghetti, as i figure out how to organize things but also UNITY don't work as advertised, remember not reinventing the wheel? unity has custom render texture, see! i read the manual, DOES NOT work as advertised, so I had to reimplement my own version but i was till using the structure, but I confirmed my ideas work, so I need to refactor to finish it, currenly only generating ambient occlusion, not yet GI, i need more flexible code. Guess what?

Material dmat = new Material(shader.GILMLit);
root.GetComponent<Renderer>().material = dmat;

doesn't work and blew up the whole project, when i'm out of the playmode, the asset is pink (it's not the shader or teh material, they have been used before and tested) and as lost reference to its default material, it's a function I'm used to used, worked in the spaghetti scene, and never met that problem before, i don't understand what's going on... I can't pay anyone to look over my project, i'm below poverty line, most people i ask for help question why i'm doing this GI instead of using SVOGI, which is unavailable, instead of addressing the problem. While latter example are related to unity, it's the fact that having problem with unity also lead to dealing with the problem of solving it and teh ecosystem around it, ie compound PAINS.

That's pretty typical of my experience with unity:

- my first games was trying to test unity for game dev by loosely porting the blitzengine and solving the technical issue I had wioth gimbal lock, I solved gimbal lock only to be stopped on my track because unity can only listen to a single contact point and that mean the project is dead, or I'm making something entirely different. And I started on unity 2 where it didn't have layer for collision, so I had to ALREADY wait for unity 3 to implement those.

- okay, let's forget running on wall, i'll make a fast racer akin to the boost style sonic game, but control at hi speed are sloppy and I figure out that's it's because unity input are sloppy and don't poll by default (like not at all), and you can't roll out your own polling in between frame because it's all done in update not fixed update, and i wasn't the only one to complain ... I guess I must roll out a slower game, but it's an entire different game.

- Boy doing an open world generic zelda like (old style not botw), but streaming can't work properly in unity, after searching in depth, I figure out all people just use source code access (and they aren't even making open world), still had performance issues mind you, which is beyond my reach, I'm poor. Of course I was already rolling out my own terrain system because unity's wan't up to task, as many people noted.

- Hey I can still downgrade to a visual novel like, i wasted enough time, but i'm gonna waste it on the awful UI situation now, fighting it to squeeze any innovative ideas i didn't realize they were...

SO WHEN people say unity is a generic engine, it pretty much failed at being generic too, things that comes out of the box with blitz engine, which is an obsolete engine made by someone who actually made games, aren't even available, so I can't even work around.

WHEN you see scrappy little team making it, you have to realize, like i did after research (read interview, looked at linked in, go on discord), they ALL comes from very specific socio economics background, like when you see fresh students out of expensive schools complaining about food (more than 10€ a day vs less than 3€ like me) and rent in their small apartment right in the middle of an affluent city, which they can survive thanks to the backing to their spouse who is either a doctor or lawyer, and with the disapproving backing to the family, which they leverage for networking opportunity, and whom they can fall back to by trading their dream for working in the family business. Well that's not me, there is a limit to self made man, and unity isn't enabling it, that's why a lot of unity devlog are either safe and similar project (survivor bias), or meme game for youtube cred.

I'm just deep in sunk cost fallacy now. I can't say it's unity's fault for sticking for too long on something that do not work as advertized. I'll try to finish the current project still. I'm slowly transitioning to javascript anyway, html is quite portable surprisingly, hyper performance be damned if I can finish a game, so far it's going better.

That's my salty contribution to the thread. Soldiering my way out.

 

I develop all kinds of interactive 3D applications...including games of all types. Have sold combat and navigation systems on the asset store. There is a game in my sig file captain observation. I am diminishing a fruitless idea that gets the folks who WORK daily with the application nothing of noteworthiness. Bark up another tree.

 

And if the experience doesn't suck then what has been gained? Yeah.. I have been reading the thread. It is a pointless endeavor for Unity Tech to make a game. The amount of work on art assets would chew up a huge chunk of the budget. For what purpose? To satisfy those who get pumped on a photorealistic terrain or weapon texture. What breakthrough is to be achieved except the fabled "this sucks" routine espoused on this thread..unless it doesn't. The it will be ...well it didn't suck because it wasn't full AAA with giant hordes of whatevers..but just do that and you will feel our pain... I will tell you what sucks. Creating Windows applications. Do that for a while and you will think Unity is a dream content creation app. Furthermore..folks who diss on the MIDI focus. You want a three hour score with your game. You can download a giant audio file of a few hundred megabytes or a MIDI file of a few hundred kilobytes. I can chop and channel the MIDI files to create multitudes of variants on a them..in code with just a few lines. I can speed up or slow down the score or use it to trigger soundtracks. I used to use a keyboard to get keyframe animation done in C4D.Unity used to think having MOD files, that rinky dink music from 80's computers, was the bomb. Entirely useless and only two crippled out of date apps to create MOD files with.

 

We want the engine to have things like a better workflow, better documentation, and better support for the features it does have. MIDI controller support is an absolutely minor niche feature and everyone here knows it. Stop trying to derail the thread to turn it into your personal feature requests.

 

HDRP anyone? obviously you are reading the forum wrong, the game proposed aren't in for realistic graphics, but to point at pain points like streaming, dealing with huge quantity of assets, terrain modeling, online updating, and performance at scale in a project. The theme is basically "scale" and "tool" not graphics.

 

Scale and tooling are a good cause. That can be achieved without the "pain" of making a game. I know pain, heads into walls and etc. with Unity. From a project management standpoint, this tract of thinking will not lead to getting tooling and scaling done quicker. You want scaling then scale till it breaks, find out why and then fix and scale to next threshold. Not wait for the art department, then scale a few assets thousands of times and wait for more assets to be dolled up. Nobody here knows what is being done with terrain but we have had some of the best minds take a crack at it for their asset store packages and none of them are jaw dropping IMHO. Even MegaSplat, awesome as it is does not perform texturing of terrain photorealistically. I do not think making a game is going to overcome these hurdles. MIDI is tooling. WebGL Mic Access is tooling. They are useful for both games and commercial applications and social site type interactions. FPS and BattleBot will provide tooling for FPS and BattleBots. Contract me and I can make either at scale with current Unity tools.

Now..ket's talk cash outflow for this dream game. Fifty to three hundred bucks and hour times several team members times several months or more plus crunch time pay. Let's say seven team members at 1000 hours each..somewhere around six months..That is a minimum 350K ..plus whatever work they won't complete because they are training a Unicorn to fart rainbows...or something..

 

Whatever. I am making a point and tomorrow I could care even less about MIDI. They won't add it and if they mention then some fools will come out and tell them not to waste resources on it. WebGL mic access was a nightmare that has taken three days to solve properly and added a few dozen defines and hundreds of lines of code to make work properly. I need proper cross platform tooling as advertised. So do all those folks searching for the same answers I came across when seeking info. I leave you back to your ongoing snark fest with me.

 

Why are so obsess with photo realism? anyway, this is something that is solved through virtual texturing in the AAA realm, that and big render farm to crunch it into something useable, but nobody is asking for virtual texture, and I think unity already support that so that's a non issue twice.

But yeah tooling, workflow and pipeline, when you mention the genius on the asset store, they are beholden to unity's limitation too, they can't fix unity itself. That's why the technical limitation hurt so much.

You miss the point about battlebot fps anyway, the genre isn't the focus, it's the fact it recoup all the pain point that make it a graal with unity, it's a stress test that have all the issue and concern to solve.

Also nobody is asking for a fortnite competitor, as in all assets must be world class implementation, badly design asset with 50 000 tri and all pbr textures stacks have the same technical impact that beautifully crafted zbrushed marvel. They could just procedurally generated the result to go fast and it would be similar, no need to involved too many artist. It just need to be possible, not pretty ego stroking demo like they did so far.

Unreal does it, why unity can't, their demo don't involve the army necessary to make a succesful game, it's a F***ing demo, all they need is to kit bash ready made art, unreal 5 lumen and nanite, was basically copy pasted assets from a bank with a few people, surely unity can do it. Heck even unity vanity demo that lead nowhere useful where done with a few people. Just crank the automation.

That's why people added in their proposed game, fortnite BUT only with a team of 5, 10 max. We are able to recognize what's possible or not. In fact I'm going to stress test unity too by making a 6x6km passable open world mock game, I already did a very simple first pass once, without game system, so that's doable to a point.

That's why I don't believe the argument of needing huge team of highly skill hold that much. That's now what's ask, we aren't asking for a competitor to fortnite. Possible don't mean it has to be good. The argument is that even a mediocre to bad asset flip of that isn't possible. The open game project failed on less than that.

 

That's exactly what people are asking Unity to do. Use their own tools at scale, see where it breaks, then fix it.

I agree that they don't need to make a game for that to happen. But a bunch of it hasn't happened, people don't feel listened to, so they're throwing down the gauntlet in an attempt to have Unity find out for themselves.

I agree it's not an ideal strategy, but I also understand the thinking behind it. And while I didn't read the full thread in detail, I don't recall anyone seeming to care about the resulting game, just what Unity learns while they make it.

 

We don't want Unity to make a game. We want Unity to want to make a game.

 

I understand the thinking behind it. I agree with you for the most part. I don't agree with the thread sections talking COD, Fortnite, battlebots as examples to emulate. I don't feel listened to..but I either make Unity work for the contract I am performing on or don't eat and sleep under a bridge. I have a much different mindset as to getting things done and can't afford showstoppers so i find my way around them. Networking, multiplayer, massive sims, animatronics and DMX stage control, Olympics gymnastics pose detection, natural language processing, vehicle control interfaces, fight games, racers,.. I am all over the map and got every one of those done..but I am to believe an FPS or battlebots is a pain dealer? I get nailed for focusing on the photorealistic cliche..but look at the thread complaints.. CryEngine does better terrain. Unreal's nanite and Lumen need matching.. If I got the wrong impression it was those pipe-ins that cemented such. If yer game relies on photorealistic terrain and you can't finish it because then maybe yer game mechanics suck.

The best thing about Unity is the deep community resources without which I could not have completed those projects. Seems every time i get snagged someone had a snippet of a clue how to unsnag myself.

 

I guess settling for mediocrity works too you are right.

Since you are so good, i proposed you clone yourself and sell them to any unity user facing difficulty so you can share your innate science.

Anyway, most of my experience is finding work around and still get screwed up for any ambition, and I'm not aiming for photo realism. You are settling for mediocrity is what artist should do, that makes the stomach full.

You get all the wrong part of using fortnite as an example too. For example, the guy who made limbo, they made another game i forgot the name, it's a side scroller, they had streaming issue, its not a photorealist shootbot game. The format and genre don't matter as much as the collection of challenge. They solved streaming by having access to code, obviously i don't have that cash. Obviously any game is equivalent to any other games, just make a game ffs, the experience ofvthevgane do not count, no ambition no problem! But i won't last long with unity anyway, the work around is increasingly to ditch it, as experiment outside are being conclusive and actually translate into marketable skills.

 

I expect most of them are asking Unity to make what the complainer is making. Because that's what'll expose the pains the complainer is most familiar with.

I suggested open world because I'm doing that, successfully, but oh boy Unity could make it easier, and with simple stuff that's help other games too. But that type of game forces you to deal with it, rather than working around it.

Same deal with suggesting realistic graphics. They're not always desirable, but going for it forces you to face graphics issues head on, rather than stylising around them.

But, again, I don't think this is the best way to spend effort. Even in big studios with in house engines, the engine makers usually aren't the ones directly making game features.

Perhaps taking a leaf from Atlassian's book would be an idea? Force every engineer to be in contact with customers regularly and to solve their problems. Not via a QA department. Personally. Every week.

 

Inside is the name. The authors PlayDead also migrated to Unreal for their future projects after two Unity games. Albeit, some of their issues have been somewhat addressed. They built those games before production (semi) ready custom SRP/URP and incremental GC. The streaming problem, however, is still relevant. And they had full source access, too.

 

There was a thread recently where someone asked about wheel collider speed not being writeable, and a valid reason to be able to do so (network rollback), then out of the blue a Unity employee asked if they were interested in testing a custom build with that feature. My mind was blown away. Usually there's so much red tape, so much struggle to produce a bug report that QA deems "worthy", and after that who knows how long it's going to circulate in their backlog until it draws someone's attention, that it felt I was reading a thread from an alternate timeline or something.

 

A demo made from kitbashed ready-made assets is not useful.

The point of the engine is support the pipeline all the way, and work efficiently all the way. When you skip asset creation, you skip also many bad spots in the pipeline which currently exist.

For example, currently both blender and fbx export is S*** and cannot out of the box autoconfigure materials. This stuff is time consuming.

Unreal engine also has similar problem, with all their fancy demos their PHAT editor is total trash. It is on dumpster fire level. That's probably why unreal games usually have very badly done ragdolls.

Make it team of 1. Then it will be very interesting.

Actually, let's go further.

* Start from scratch.
* Team of one.
* 48 hour limit.
* No Asset Store.

Ludum dare, pretty much. How about that?

 

Mh, rather bad idea. Then the taking from all this is extremely dependent on the coincidence on how skilled and passionate of a person that one dev is.
"small team of 10" sounded reasonable and makes it more feasible to reach the scope of a game where most painpoints lie.

Eh what? Really don't think Game Jams are Unity's weakpoints. Rather the opposite is the case.

 

This is pure hogwash son. The innate science is determination and good search engine skillz and twelve years of daily doing it... for pay or not eat.. My clients certainly do not think I am mediocre. There is always more than one way to skin a cat. SpaceX and Magna are looking at using the technology I am currently contracted for.Northrop Grumman already built a 100 million dollar business unit out of this tech. So..your snark is noted and i consider it unbecoming of you..but go for it if it cranks your turnstile.

 

A team of ten for six months is a half million dollar investment...minimum. To prove what?? That is alot of cash..not stock value.. Unity need to up their game..not make a game. I had a physics bug for three months and produced over 100 bug reports and kept getting BS replies from diversity hires. Finally one of the Senior PC platform engineers spotted my forum post, went and checked my bug reports and apologized publicly. Three days later the bug was gone. I think the platform engineers should be checking forums and proactively handling issues that are showstoppers. There are people that care about there work at Unity.. These folks are not recent hires..who seem to be enjoying a paycheque as their work domain. They had a Bugz Report guy who insisted my issue could be resolved by updating DX 11..umm you can't..

 

I honestly don't think so. One person will have unbalanced skillset. Meaning they'll likely be good at one thing and worse at another. And if the engine can accommodate all such users, that will be good for the engine. Wasn't the idea to democratize the game development (long time ago, in a galaxy far away...?)

 

@ippdev I hope your big impressive clients are providing health insurance because I think you're in danger of tearing a rotator cuff with all this back patting.

But hey, what can I say. When you are smarter than everybody else, the only sensible thing to do is brag about it constantly.

 

You should talk BigTime Master..LOL..
edit...went and checked your website for skillz. Seems the images are broken or don't load. Ya ought not to be tossing stones till yer glass walls get replaced.

 

That's just very narrow though. Yeah if everything works out as we hope, then the engine accomodates that single persons skillset better. But not the other 50 combinations of skillsets out there...
Also it's kinda reasonable of Unity to not focus that much on optimizing for projects of One-man-shows. Very rarely those projects blow up so much to reach the pro-licence threshold where Unity starts to get any revenue...

 

Yup. And, it should be noted that Keijiro has several working MIDI projects for Unity:

https://github.com/keijiro/Minis
https://github.com/keijiro/MidiJack
https://github.com/keijiro/MidiAnimationExample

I've had zero issues using MIDI with Unity. I do agree with @ippdev that Unity making a game is not going to be that helpful. Unity is a generic framework that doesn't push you into making a specific type of game and as Keijiro shows it is generic enough to build MIDI input on top of. Problem solved!

 

Blomkamp even made some robot-apocalypse shorts with Unity. Blomkamp-inspired mech, the multiplayer game might awesome direction for the Unity game. Especially such an ambitious project would rape every single incorrect assumption of Unity development. It would break apart in the first months, but all of us would benefit from that in the long-term

I just suspect that CEOs there are too afraid of tacking such classic development storm

 

KLewij

Keijiro is an amazing developer who shows just how far you can push the engine in many fields. I think http://unitylist.com shows the breadth and scope of applications that can be serviced using the Unity framework.

 

The percentage of folks using Unity to make MP massive battle games is very tiny. I would think that considering it's the tool of choice for many other high end industries the CEO ain't skeered of your proposal. He is looking to service the other industries rapidly climbing aboard including but not limited to interfaces for vehicles, military simulators, industrial design in VR, AEC industries vast needs, film and cinema, animated features and series, visual interfaces for an entire range of AR and VR, animatronics control, stage lighting control, VJ toolsets, medical imaging, surgical training and remote surgery, educational and edutainment applications, industrial systems and safety training. Doing this will keep the stock bouyant, investors happy, money coming in the door and the ability to hire better employees specific to issues. They also did a huge hiring wave which will need to filter out via the Pareto principle.

 

Asset creation isn't skipped, it's already done with readyb made asset, problem that happen in blender are blender's juridiction and should be reported to them, it's not unity's job to do blender's job.

What's pure hogwash is that you assume any other people on this thread don't have mad leet haxor google skillz. By the time I was done with an issue with the render pipeline I didn't complain about, I learnt the entire technical history of occlusion, rasterization and game rendering engine, can probably implement S and C buffer, understand the difference between bresenham based triangle drawing and the more simd friendly block based version. I'm so paranoid I go up to the 10 pages so i'm sure i didn't miss the obscure wayback machine link in a non popular but very technical forums.

I mean we can dick wave all the way, the problem doesn't go away:

LOOKS LIKE EVEN YOU CAN'T SOLVE EVERYTHING WITH LEET GOOGLE SKILLZ

 

It's funny that someone as technically knowledgeable as you isn't able to see how a test case to handle massive workload over online infrastructure can benefit from all these industry. I mean Unreal demo where game like, yet cinematographer DID see the benefit right away, so did archviz people.

Some aspect of the engine are non negotiable, if they are broken, why use the engine in the first place, fine I can do my own GI (even if you think it's not a necessary part of a good whatever), probably can roll out a rudimentary physics, culling system, scenes management, etc (which I do constantly) ... but that's assuming unity's issue don't stand in the way like they always do in unexpected place. When basic stuff like file access (streaming), rendering or inputs can't be work around because the engine take authority on them, you are screwed. That's what we want them to test without messing with source access. They already failed at the simplest game (the open game project) they have work to do.

 

Just ignore ippdev. He basically only ever shows up in threads to tell people how important he is in roundabout ways and present himself as part of some sort of silent majority of Unity users who all need extremely niche features to be added to the engine.

 

Typical disinformation from @Murgilod.

I've been using Unity to develop state of the art systems for lockheed martin to defend against the jew space lazers. If unity was broken, you'd all be dead so you can thank me later.

Also, nobody knows anything I don't, so why is anybody even posting?

 

Unity, are you going to make a game or not ?
Please make a statement or close this thread before your users gouge each others eyeballs off.