Hacker News Reader

Comment by hesdeadjim | Hacker News Reader

OSS isn’t tenable as a competitor to any real engine you want to ship big games on. Just the QA testing budget alone that is required to support all consoles, PC/Mac, iOS and Android outstrips the budget of even Godot.

What we need is a VC backed, for profit engine, that is up front about taking 2% of revenue after $1mm for perpetuity. And only being an engine company. Period. And like Unreal, source available, just not forkable for commercialization.

There is immense room to eat Unity’s lunch and capture its entire market share. Especially when you consider that a new engine can discard 15 years of legacy support Unity’s team fights against endlessly.

Edit: downvote me, but I’ve been doing this for 15 years and this is the only way. blender is not a useful comparison, it supports two platforms and is built on other OSS software that is GPL.

View on HN

I don’t think VC-backed engines can avoid enshittification. Unity is really demonstrating that now. OSS is the only way.

Depends on who the backers are I guess. Perhaps not the classic kind — like say, convince MS, Apple, and Sony to pony up $10mm each. It’s a good play for them to not be caught in this Unity fiasco and $10mm is a rounding error.

> $10mm is a rounding error

And you wont manage to achieve anything with that. Maybe Unity could spinoff its ad business (together with some major liabilities like Ricitelo and other execs) and then MS could buy the engine part.

$10mm each. $30mm backing and you could get a pretty solid tech base to build a game on. Especially since there are some great new libs out there for things like physics. The Horizon Zero Dawn lead programmer has an OSS MIT engine that is very advanced and less bloated than PhysX.

> $30mm backing

Realistically that would still only be enough to get something you can make a cool looking demo and that only run on desktop.

Building a general purpose engine that is suitable for many different workflows and run on pretty much every singe platform like Unity is very trick and takes a lot of time (and trial an error). Dev UX, APIs, platform integration, QA, all the third party packages and infrastructure are much harder to solve than shiny graphics or advanced physics. Whatever you come up with that $30mm is only likely to cover a small subset of use cases.

v1 you would not attempt to be the be-all-end-all. Target iOS and Android since those are core Unity markets and throw on PC+Switch as the secondaries.

$30mm is more than enough when you ditch legacy stuff like anything that isn't a Vulkan/DX11/Metal renderer.

> $30mm is more than enough

With the right team it might. Realistically they'd just under promise and undeliver (or not deliver at all).

> stuff like anything that isn't a Vulkan/DX11/Metal renderer.

That's also not the complicated part.

Traditional VC companies wouldn't work because they'll always enshittify. Even on consumer space, you can see enshittification from MS, Sony and Apple. I really don't see a reason why the companies you listed would need that 2% royalty at all honestly. They already get the 30% from the sales themselves, and could very well financially support the existence of an OSS game engine with no extra royalties. Sony and MS already develop the Unity and Unreal integration for their platforms without charging extra royalties.

The devs using the OSS engine also wouldn't be stopped from spending resources on making the engine better for their uses (and in turn, making it also better for everyone), similarly how it works with Linux.

Yeah it seems like enshitification during the long tail is the point of the VC model across all industries. Give product away for free, figure out how to screw the install base later on.

Can't blame the VCs for trying this model, but they've played society for suckers. We let them get away with this for too long, becoming dependent on their "free" offerings.

At this point I'd be willing to get rid of my free stuff altogether and pay a subscription for products I like/use as it would incentivize the companies to add features that benefit me.

For example, search on the web is dying. I would absolutely pay money to replace Google with a search engine I trust and that works for me.

Try kagi.com. I'm a very happy customer myself.

Maybe I’m an outlier, but I blow right past their limits on any reasonably-priced tier. I use my search bar for all sorts of inquiries, serious and frivolous, including basic addition and subtraction. I tried it for a month and found that my usage far exceeded what I was willing to pay, which was about $10 a month. If they can get the costs down, I’d happily subscribe, I think the results were very good. Google’s results have become terrible over the years, it’s hard to fathom how it’s still their cash cow when the product is practically unusable.

Search is dying. The quality of search I mean, which doesn't bode well as civilization pretty much runs based on our ability to search the web for commercial/ government / political work.

It's a problem. Whoever control search will control what we see hear and read. Dangerous times. Hoping that somebody can build a subscription search that works and returns real competitive-ness to the information marketplace

> Yeah it seems like enshitification during the long tail is the point of the VC model across all industries. Give product away for free, figure out how to screw the install base later on.

It needs two sides for this to work:

1. a company attempting to "fish" customers with this approach

2. customers who fall for this trap, and did not create an emergency/backup plan when the respective company becomes "evil"/untenable.

Every company sooner or later becomes "evil" (and be it because of ownership changes because the original founders die or want to retire). Having made no "backup plan" for such a situation to happen is in my opinion gross negligence.

Even OSS can be enshittified. Nothing is free from it unless it is legally barred.

VC backed for profit engine is precisely what Unity is though?

So is Unreal and CryEngine.

Both of which only survive on graphical fidelity, which is not what Unity was touted for.

Unity was for filling that small-to-medium sized games in terms of man power, if you got 15 years of experience then you should know that.

I think it would be more interesting to see a Coop backed engine, where the COOP has to listen to it's customers.

Since my biggest annoyance with proprietary software or in this case game engine is that you can create the most elaborate sub-engine (for example a fully-dynamic RPG) and see it all come crashing down because some dimwit at Epic Games reeeeally wants to push the latest eye candy that breaks your sub-engine.

That's the one thing OSS is far superior at, you can fork the engine form a new developer community or even work with the OG maintainers to incorporate the sub-engine as a module and as such be officially supported both by your team and theirs.

Meanwhile with proprietary software, you're stuck at this particular version with no legal way of ensure the survival of the project, RPGMaker games are the perfect example of this, as soon as you venture past it's originally intended fidelity, you're essentially going to be limited by the engine in terms of performance.

> Since my biggest annoyance with proprietary software or in this case game engine is that you can create the most elaborate sub-engine (for example a fully-dynamic RPG) and see it all come crashing down because some dimwit at Epic Games reeeeally wants to push the latest eye candy that breaks your sub-engine.

Uf, feel that one. If the engine isn't designed with your genre in mind you may be making some significant structural changes(which will be interesting to see how open source engines handle that).

Even if they don't take the engine a different direction the up-leveling cost of getting new features/bug fixes is significant. Back when I was working in industry we had one person who's almost fulltime job was to merge -> fix compile errors -> fix runtime issues -> checkin change and uncover more edge cases that got broken -> repeat process, it was brutal.

>Even if they don't take the engine a different direction the up-leveling cost of getting new features/bug fixes is significant. Back when I was working in industry we had one person who's almost fulltime job was to merge -> fix compile errors -> fix runtime issues -> checkin change and uncover more edge cases that got broken -> repeat process, it was brutal.

And here I thought industry is that one programming market with no awful grinds. I do sympathize though, webdev frontend+backend is just hellish (but you do have some tools to alleviate the pain, but when you have to rewrite the app 6+ times it really breaks you).

Well most games usually get built once on one version so it doesn’t matter if the engine changes.

That used to be a lot more true back when game studios treated a finished game as almost immutable.

These days it's quite common to see new patches and even new content for single player games.

Anyone down voting this needs to pause and take a good long look at their cloud provider bill.

Why are you not self hosting?

Probably, its because the its very convenient having some else doing the release cycle QA, deployments, infra work for you.

The same goes for game engines.

Are you going to start self hosting tomorrow if AWS announces it’s charging you $0.2 per db query on top of the existing pricing?

No, you’ll swap cloud provider.

Maybe some people will try self hosting… but they’ll find out it’s actually more expensive and harder than it looks, once you start doing it.

Anyone telling people to go self hosted in this situation would, rightly, be giving bad, naive advice.

So, maybe have a moment to think about that, before you down vote someone for saying “that’s bad advice” to use an open source game engine.

Exactly! I, up until recently, have been happy to pay Unity six figures for enterprise support for exactly this reason.

I even have engine source, and guess what, it’s wildly complicated. My team doesn’t remotely have the bandwidth to do anything meaningful with it even if we had the right to make modifications.

The argument for open source always neglects this reality. Just because I could change something doesn’t mean I have the economic reality to actually change it.

>> blender [...] supports two platforms

Windows, Mac and Linux; that's three, right? And it supported many Unix variants while their marketshare was significant. Where does your "two platforms " come from?

>> I’ve been doing this for 15 years

Most commenters here don't have 15 years of gamedev experience, but many here watched Blender turn into a tool among the leaders coming from a "very weird UI" in about that same amount of time. Why can't the same happen to Godot?

>> PC/Mac, iOS and Android outstrips the budget of even Godot.

We heard the same about GNU, GCC, Linux, Apache... look back at what happened to most closed source Unixes of the 90's. Yes, game engines are a different market but your argument is the same we heard other times.

Blender has been around for 29 years, to be fair it’s only been open source for 21, but still it took an extremely long time to get to a point that was comparable to other tools. It’s still definitely not as used as Maya. I do genuinely feel Godot can be adopted widely because it’s better than any other open source game engine that has been around, but blender took an absurdly long time to be competitive.

> 2% of revenue after $1mm for perpetuity

I'm sure Unity would be fine with that. I don't see how could they hope to collect more than 0.5-2% if averaged out over their all customers even with the new pricing.

I guess the problem is that revenue share is very hard to enforce, by adopting something like that you'll be probably giving up on a significant proportion (if not the majority) of fees you could get from developer in China (and some other countries in similar situations). Install fee (while a bizarre idea) should be easier to enforce.

I don't understand what QA budget has to do with game engine budget and why the latter needs to be larger. What's the budget of Linux again compared to stuff that runs on top of it?

First off, you can’t OSS consoles. The manufacturers require an NDA and arpproval to gain access to the hardware and SDKs. This costs real money.

Second, every console and to a lesser degree mobile are constantly changing their SDKs and that requires a lot of testing to ensure that your engine works across a huge permutation of platforms. Throw in the need to support an LTS version of your engine and it balloons in cost.

> First off, you can’t OSS consoles.

All those SDL games already released on consoles and Godot soon gaining support via W4 beg to differ.

Yes, it's pointlessly hard and annoying because of platform stewards and all the hoops they make you jump through, but it's certainly not impossible nor infeasible.

I’m not a game developer nor do I plan to become one but the thing that seems strange to me is all of this talk of revenue sharing with the development tools. If I create a successful app or website using Visual Studio, Microsoft isn’t going to take a percentage of my revenue. Even Apple won’t take a percentage to use Xcode and only charges to publish on their platform. Why is game development different?

Microsoft and Apple get their cut by charging 30% on their game platforms.

I know. So if you use unity to create a game that is being published on iOS or the Microsoft game platform, are you not being charged 30% from Apple/Microsoft and then a per install charge from Unity. That doesn’t sound sustainable whether you are a small or large developer.

Because that is the business model that Epic and Unity figured out for their engines, and poor game developers saw as preferable than paying once upfront, which might or might not have been a mistake.

You are also not shipping 5-7 digit lines of code from the dev tool in your project.

I spent some time in the trenches on this, and concluded what is needed is some sort of agreed module/integration mechanism and an ecosystem of open and proprietary modules. (Quite similar to how many large game devs operate already).

No “open engine” will ever be enough, by definition, as the games people will pay for will be those which were hard to make, and tech advantage is one well proven way to do that.

The tech industry is beginning to learn that barriers to entry, such as those Nintendo created in the 1980s or existed in the early days of the App Store, actually helped the market function to create much better results. Apparently Atari will be repeated again and again.

This is compelling to me in general. Getting a core engine stood up that is capable of this flexibility would be a hell of a challenge, but a fun one!

What? VC-backed is all about the strategy Unity is employing: capture the market and enshittify afterwards. They literally can't "be an engine company. Period".

Unity provides a good example of why it would be a wise strategy. You don’t need a unicorn in this space, just something that can throw off serious cash over time as that 2.5% compounds.

Ah yes, the only way to make fun and successful games is to introduce corporate greed. History has proven your 15 years of experience wrong many times.

I was about to agree with you. But then I thought about it. Linux kernel handles way more so it's definitely not outright impossible.

Linux kernel doesn’t require signed NDA contracts with console manufacturers.

In exchange for what? What does that give Unity access to which Godot doesn’t have access to?

In exchange to access the sdk and hardware. OSS Godot doesn’t have access. Only their newer for-profit arm does, so now… not OSS?

I'm not sure what you are saying here. Having a "for-profit" arm does not exclude a piece of software from being OSS.

Doesn't source available go against "support all consoles" as the SDK of these platforms are under NDA?

Also Blender officially support more than two platforms but I don't know how it's in topic here.

Source available for consoles would be gated on your company having the same NDA.

So okay it doesn't really matter if the engine is open source or source available, I see.

Console port support for Godot would rely on commercial/proprietary offerings due to the fact that you can’t open source those code (as you have to incorporate vendor code).

So that aspect would not be OSS.

> OSS isn’t tenable as a competitor to any real engine you want to ship big games on.

Indie and big games in the way you mean are pretty much mutually eclusive.

Wrong. My indie studio is “small”, but we ship an open world game that is only tenable on Unity because our our minspec. I’d switch something not Unity ASAP if something comparable entered the scene.

"Indie," or indepedent, just refers to the studio. Indie studios can work on ambitious titles.

Subnautica is a good Unity example, but there are others

As I have been pointing out for years, and this clusterfuck shows yet again, if you rely on any platform, you are not indie.

Some people also say that indie also means small, otherwise the most indie studios out there would be the likes of Valve and (Microsoft-)Activision-Blizzard.

Independent, i.e. not working with a large publisher, like Microsoft, Sony, etc. That precludes those publishers themselves, of course!

Why would that preclude those publishers, if they also own studios ? (See also : Paradox.)

Also, Valve and Blizzard aren't (weren't ??) publishers.

Valve seems like the only company that would do this reasonably. but they've got better money makers to deal with

"VC-backed" seems like a naive opinion at best. Also, doesn't UE already do everything you describe? Why is another one needed?

Half the Unity games are complete and utter garbage anyway. Most people don't need the complexity of a full-blown engine for the kinds of games they put out there.

QA costs could ultimately be paid off by people who actually use the engine, no? If they contribute back or, at least, create an issue things could work out.

You can’t risk your business on broken software. An engine company also has to support old releases of the engine for a certain period of time because you can’t just upgrade your shipped game to every new version of an engine that might contain back compat breaking changes — or changes that have too high a cost to port to.

You'll have a ratio of thousands of users to one contributor. Even less if you count useful contributions. Game engines are complex beasts.

Source: I have an OSS project that's vastly less complex than a game engine and worked on dozens of other OSS projects.

Reply on news.ycombinator.com