I don’t have the answer as to why, but perhaps, because a lot of game customers NEED to see their work in DirectX. Re-use their HLSL shaders and plug their DX setup into Maya regardless of what OpenGL can or cannot do.
It is not that Maya is suddenly only DirectX. So it wasn’t a choice between DX or OpenGL.
Just that the DirectX pipeline got some love.
Yes, DirectX is a Windows only feature. What’s your point?
@Dutch: I don’t see how that’s lame on Autodesk’s part. DirectX is a Windows/Xbox only API. Nothing Autodesk can do about that. No reason to spite them for implementing a feature that the majority game development users will benefit from.
I find it lame in the sense that they could have opted to work on many, many other things that aren’t platform specific. Finish VP2.0 for example, before starting on yet another offspring project. That’s my opinion and I’m sticking to it.
That is a requirement of DirectX, not an Autodesk created requirement. There’s nothing they can do about supporting DirectX on other platforms.
Well, considering Softimage has had DX and OpenGL shader support in it for the longest time, I would think they could utilize a lot of the experience from when that was implemented, to make this implementation happen a lot smoother and easier, regardless of the fact they are different applications.
Whose to say they aren’t improving VP 2.0? We’ve already heard news about work being done on it. This is just a SAP update, not the next version of Maya. They’ll probably support OpenGL shaders in the future as well, it’s just a matter of what’s most important, which would be DX by a landslide. If they implemented OpenGL instead of DX first, DX is still needed/wanted, due to majority game industry penetration. It’s not one choice being better than the other, but both choices being needed and choosing which one is more important to get implemented.
Studios are their biggest market, not freelancers with Macbooks or a dedication to OpenGL.
This article is about a 3 year old game, that has now been ported and had a lot of optimization done to it. It has little factual merit to it, since it’s working on the assumption that the DirectX version was perfectly optimized. And, my good man John Carmack tends to disagree with that as well. He constantly states that he believes DirectX to be better than OpenGL these days. He even said it in his QuakeCon keynote last week. And I’d trust his word on such things a lot more than I would Valve’s Linux programming team.
You’d hope so. Because it’s not finished. Even after what… 2-3 years… there’s still a lot that doesn’t work with it. I can’t use it at all in my pipeline because it can’t display many of the geometries and shader networks that I heavily rely on.
You are conveniently forgetting the VFX industry. Where by and large the main choice of software is Maya on Linux.
I’m not one to begrudge my gaming artist brethren their tools, but this is one feature that would have been better left to 3rd party developers in my opinion. Or at the very least put on the back burner until such time that VP2.0 can be considered “complete”.
The nature of adding features means that developers will always be prioritizing one set of users’ needs over another. In this case, it seems infinitely more productive to look at the DX11 viewport as a feature intended to support game studios (just as nHair, for example, was arguably more VFX-friendly, or at least not for game devs). The limited OS availability is a byproduct of the target audience, not a feature goal itself.
The comparison to nHair is flawed. nHair is a feature directly accessible to all Maya users regardless of their choice of platform. Wether a technology is useful to an artist or not is another matter entirely and is not the point of contention here.
Furthermore, ‘infinitely more productive’? A bit extreme don’t you think?
Excellent point. That article is pretty light on specifics of which versions of each API they used in their test. I just wanted to point out that perhaps OpenGL will have some heavy hitters in their corner soon. Healthy competition is always a good thing. I’d certainly like to see an unbiased comparisons of the latest versions though. That being said, it doesn’t seem right that Mac and Linux licenses cost the same as a Windows license, even though the feature set is not entirely cross-platform.