potato_hamster said:
Great. Nintendo joined Khronos. What exactly does that mean for Nintendo? Does it mean they have unfettered access to everything Khronos creates, for them to use how they see fit? Probably not. That's not how it works. But if it did, you're assuming the Kronos API as it stands does exactly what Nintendo needs it to do, and has a resource footprint that meets the demands of Nintendo. Sure it might be low-level, but it might not be low enough, and it might be nowhere near the level Nintendo needs it to be. I can almost guarantee it does not meet Nintendo's needs as it stands, and Nintendo will have to modify it dramatically to meet their needs. That's if they're even using this technology in the NX. They might have just joined the group to evaluate the technology. You should see all the groups Sony and Microsot is a part of. It doesn't necessarily mean anything. And FYI, it is the API that is "powerful" in the sense that any piece of software can be more powerful than similar software. Some APIs that are more capable and have more functionality could be considered to be more "powerful" than others. This should be obvious. Don't argue over semantics. Also you're failing to understand the fundamental issue that creating the infrasture ( the APIs, the development kits, the developer tools etc.) is extremely difficult, and you continuously trivialize this. Making all of this work they way you picture it is an very very very difficullt task. Of course this is theortically possible, but is isn't practical - especially not for Nintendo as I've outlined above. And you just kinda shurg and say "Nintendo will benefit by doing so so they'll just do it". They have to actually be able to execute this! You keep bringing up PCs as if that really matters, and I'm not sure why. Sure PC games run on thousands of different hardware configurations, using engines running on APIs and OSs that are vastly different than what is used on consoles. You can deny that all you want but those are still the facts. Console OSs and APIS are still take a fraction of the system resources PC OSes and APIs do, so it is not the same as comparing apples to apples. Optimizing for one architecture, and one OS might much more efficient than having to do all of that for 2 separate platforms, but optimizing for one architecture and one OS, and one specification is much more effiicent than optimizing for two similar archtectures, two similar OSs, and two similar specifications. The resources required to do so might be magnitudes bigger than optimizing for a single specification - don't forget that.
It is not easy to implement the solution you continue to brush off as trivial no matter what the huge benefit is. Sure the upside is huge, but only if you make it happen, that that if, no matter how much you deny it, is a huge one. |
Considering that Vulkan is open source actually it means that anyone can use it and that includes Nintendo.
Yes Nintendo would get unfettered access to it. Vulkan is everything a developer could want in the way of an API, Low-Level, while also allowing for multi-core, Asynchronous/GPU Compute and it's supported by every major vendor, including AMD (the likely creator of NX's CPU and GPU Core Architecture). The performance of Vulkan in demos speaks for itself.
LOL no, an API is just a tool, now it can be more efficient or less, but in of itself it doesn't have power, because it's the hardware that crunches the numbers. If anything it's more how it's used that detemerines it's level of power. I'll argue over anything I choose, if I feel someone's being untruthful, you have no right to bark orders, like your the one in power in this debate LMFAO. There are no semantics here, you're wrong about this point.
LOL you're arguing over nothing, making new infrastructure isn't a problem, platform holders do it every generation and largely it's a matter of iteration and learning from the mistakes of past generation, or in the case of APIs, etc you can use a product that already exists and is freely available for you or anyone else to use. Hell being a contributor and Vulkan itself being open source means that anyone, be it a company or an independent programmer can re-write it or parts of it as they see fit, that's how it works with open source software.
You're acting like making tools for a single architecture is more complicated than it really is.
The hypothetical NX platform wouldn't have 2 separate architectures or 2 separate Operating Systems, one OS, one overall architecture, so all code runs natively on it, without issue.
It's so much more straight forward than making 2 separate architectures and 2 OS. You haven't even brought up a single point as to why this isn't true, you just keep on inventing stuff to try and support this fake point you keep on attempting to make.
Seriously what issue do you have with Nintendo taking this route? Because none of the things you keep insisting are true are.
Of course the PC example matters, because it's an outright example of a platform that has potentially infinite combinations of hardware, even with different architecture (because of different vendors making different types of core processing tech) and provided that Vendors update their drivers to incorporate new games those games work seamlessly.
The only difference between an API used on a PC and console is the language, because Sony and MS make them for their consoles, but that doesn't mean a platform holder can't use Vulkan or some other API that is also used on PC.
An API will work on any platform that it's been written to work on. Modern PC APIs are just as Low Level as Console ones, they do the exact same things as console APIs.
The whole point about this hypothetical NX is that it would use one OS, you ignoring this point doesn't change that fundemental point.
No resources needed don't get larger, because your making software for one architecture, the only difference between each device is power, the weaker platform can still run the code natively, it just runs less of it.
The engine doesn't have to translate a thing, because it's working on one architecture.
No multiplatform games made for PS4 and XBox One aren't practically identical, because they have different Operating Systems, they have different core languages made specifically by each platform holder, because the APIs are fundementally different, which is why games wouldn't run natively on each system.
This is why game developers have to port between each device or make different versions of games in tandem with the PC versions of titles (if a PC version is being developed). Your PS4 and XBox One example makes no sense in comparison to how NX platforms would work in this discussion we're having.
The reason NX handheld and console just run the same code is because it's written for the same API, that runs on the same architecture, hell even if the architecture was different it wouldn't matter, becaus the API would be written to work natively on both. Developers don't have to do anything, besides pick the settings they need to make it run on the weaker system.
It's just as easy as I'm saying, you inventing issues where there aren't any doesn't change that fact.
Ultimately making a single OS that incorporates two pieces of hardware is much simpler than making two Operating Systems, with two separate pieces of hardware. Each generation you make a new API, OS, etc anyway or you iterate on past technology and software.
