Hacker News Reader

You know what can translate into success for developers? Supporting Vulkan, instead of pushing unportable DX12.

The days of the average developer writing directly to a low level 3D graphics API are long over, so this is mostly irrelevant. Even so, the funny thing is it's likely that once Vulkan ships, writing a Vulkan only Windows game will be viable as long as AMD, Nvidia and Intel have decent Vulkan drivers even without Microsoft support. Using Vulkan on Mac might be impossible, as Apple is pushing their own proprietary API, Metal, on Mac OS X.

> writing directly to a low level 3D graphics API are long over, so this is mostly irrelevant.

Not really, and it depends on studios and developers. Firstly, engine developers can't avoid dealing with those APIs. And dealing with multiple increases their burden and is simply a pointless multiplication of effort which could be avoided. Secondly, some studios can afford developing their own engines, so it's not irrelevant even if we aren't talking about commonly licensed engines like Unreal, Cry or Unity.

> the funny thing is it's likely that once Vulkan ships, writing a Vulkan only Windows game will be viable as long as AMD, Nvidia and Intel have decent Vulkan drivers even without Microsoft support.

In the context of the interview, I doubt Vulkan will be available on Xbox and MS mobile devices, unless MS will suddenly rethink their lock-in attitude. In the best case scenario someone will create a Vulkan → DX12 translation layer (similar to ANGLE), since actual APIs aren't that different functionally.

In the context of the interview, I doubt Vulkan will be available on PS4 and Android mobile devices, unless Sony will suddenly rethink their lock-in attitude.

In the context of the interview, I doubt Vulkan will be available on WiiU and 3DS mobile devices , unless Nintendo will suddenly rethink their lock-in attitude.

It just doesn't matter when game engines are available and writing a plugin render APIs is just like CS 101 graphics programming, being done since Atari brought games into homes.

> It just doesn't matter when game engines are available

They don't grow magically out of nowhere. To make them available, someone has to develop them for each target platform. And multiplying APIs in the process doesn't come for free.

That how the lucrative industry of porting games came out to the in the mid-80's and it won't go away.

Succeeding in the games industry means thinking like the Demoscene culture, not FOSS culture. I got to learn that too late.

> You know what can translate into success for developers? Supporting Vulkan, instead of pushing unportable DX12.

Will it really?

> Reduce lock-in, support better cross platform development, and more developers will come to Windows

This is exactly what they're doing.

> This is exactly what they're doing.

I would agree to that when I'll see Vulkan on Xbox and Windows Phone.

Vulkan is due to be standardized later this year. Isn't competition here good? We've gone through the open source route before with OpenGL, and the result was an API that went its own way, ignored new developments in graphics hardware for years, and had a fractured API with hardware-specific extensions only finally unified in recent versions.

So I suppose AMD, Apple, and Microsoft should have held off on pushing "unportable" Mantle, Metal, and DX12?

> Vulkan is due to be standardized later this year. Isn't competition here good?

Competition is good, when it exists. But I'm talking about bad cases, where competition is artificially prevented. I.e. sure, on Windows Vulkan and DX12 can compete. But on Xbox and Windows mobile - not really, since MS won't allow Vulkan there. They aren't even part of the Vulkan Khronos group as far as I know.

I really don't get why you keep forgetting about all the other vendors that also don't care about OpenGL.

Other vendors are part of the Vulkan working group.

With an history of never supporting OpenGL.

Nintendo, Sony and Sega never had an OpenGL 100% compliant API.

What Sony had on PS3 was OpenGL ES 1.0 with Nvidias Cg and even that was barely used by game studios.

Being a logo on OpenGL website doesn't mean anything.

"Actions speak louder than words"

Agreed, actions will be needed. But if they aren't stupid, they'll benefit from using the common API.

> Reduce lock-in, support better cross platform development, and more developers will come to Windows. Lock-in isn't going to fly as something that developers will appreciate these days.

Uh, Windows has something like 95% market share in the market where developers care about Vulkan vs DirectX 12. I think Microsoft will have no problem attracting games developers to Windows, considering they're already there.

> You know what can translate into success for developers? Supporting Vulkan, instead of pushing unportable DX12.

Ah! You mean the Vulkan support that Apple forgot to talk about at WWDC, while updating their documentation to refer to OpenGL as legacy?

If Sony, Microsoft, Nintendo, Apple don't support Vulkan, the APIs is DOA, regardless of what OpenGL fanboys might think of it.

In the age of middleware game engines, the low level APIs are not that relevant any more.

> In the age of middleware game engines, the low level APIs are not that relevant any more.

I already answered you. Someone makes those engines, and those who use them would bear the overhead of those who develop them wasting their time on duplicating work.