Khronos Group Launches Open VR Standards Initiative

[kcarbotte]

The Khronos Group, the technology industry consortium responsible for the OpenGL, WebGL, and most recently, Vulkan APIs, is setting its sights on the virtual reality industry for its next collaborative project.

Khronos Group Launches Open VR Standards Initiative : Read more

 

[AndrewJacksonZA]

"Qualcomm is curiously absent from the list of supporting companies... Khronos isn’t known for pumping out standards in the blink of an eye." I think you just gave the reason why.

One of the things that I've learnt over the years is the Khronos seems to have committees for committees that are about committees and that things have seemed to take forever with them.

 

[bit_user]

Seems like every name involving "Open" and "VR" is already in use. Perhaps they'll follow the path set by Vulkan, and pick a name that's just an unrelated or invented word.

For a developer to support SteamVR (OpenVR), Oculus (OVR), and OSVR, it has a lot of work to do.

You omitted Google's Daydream & Samsung's GearVR platforms - I'm sure the new standard will address mobile. Plus, I'm pretty sure MS has an API, since they're working with HW manufacturers to create their own VR platform.

Qualcomm ... recently made the move to build a processor specifically meant for standalone VR devices.

It's misleading to call it a processor. It's just a variation on their smartphone SoCs that's better-oriented towards VR. It's not on the level of what MS did in Hololens, where they basically created a custom engine for AR.

Anyway, you'd expect them to be at the table, but I think the risk for them of not being involved at this stage is a lot lower than for the HMD and controller manufacturers.

The group is currently in the “initial exploratory phase” trying to “define the standard’s scope and key objectives.”

Right, and this is actually the big question. Will the new standard really allow a single app to be targeted at the full range of VR hardware, with the full range of tracking capabilities & controllers? That's potentially a far more ambitious goal than simply unifying differences between a few APIs.

 

[bit_user]

One of the things that I've learnt over the years is the Khronos seems to have committees for committees that are about committees and that things have seemed to take forever with them.

I can't speak to the first point, but standards aren't something that should be rushed. Furthermore, they will always lag what industry leaders are doing with proprietary APIs, but that's not to say they aren't important or worthwhile.

I wouldn't even expect to see a provisional spec from them for at least a year. Furthermore, I think it's likely they'll decompose the problem into a family of related standards.

BTW, as far as standards go, Khronos is actually pretty fast. Just compare them to W3C, ITU, or any number of other standards bodies. How long was HTML 5 in development? Or take Javascript 5, H.265, 802.11n, PCIe 4.0, DDR4, ... I could go on.