Fool me, this is real good shit ! [the VR thread]

[gumboot]

Interestingly, I've seen the outcome of a debate about whether or not X11 could feasibly be used in an embedded application, and despite everybody's assumption (that it's a big smelly pile of desktop inefficiency) it was successfully beaten into fast start-up and low overheads in time and space.

But then, I suppose I'm being kind of old-school by considering that "desktop". These days there are so many additional layers of shit piled on top by modern distributions, most of which ignore the core functionality of X11 and do everything by hand so as to make X11 as ineffective as possible, that X11 probably is considered "low-level" in the modern world.

[fog]

The Verge's Hololens "it's really great except when it's really terrible" vid is worth watching.

[pogo]

Thanks for that: tried looking for a proper report and that’s better than anything I found. Confirms my own suspicions that this is very far from the finished product in a number of ways. It will be years before Microsoft’s ambition (that earlier footage: “basically a lie”) matches reality.

[Whiteball]

Look for the shit, you'll find the shit. Look for the gold however and that's what you find....There was a lot of positives in that. Certainly more positives than negatives thats for sure.

Nothing is perfect yet in VR or AR but Hololens is very good and getting their faster than others....

[pogo]

We’ll see. I think it will only be worth taking seriously when developers have their hands on it. That will mean they have something reliable, something that works away from tightly controlled demos, something close enough to final to be worth working with. It then might still be years away from release but at least release will be on the horizon.

[gravy]

So only 'the best screen you've ever seen'

Hope they buck up their ideas and sort this shit out!

[gumboot]

So only 'the best screen you've ever seen'

Hope they buck up their ideas and sort this shit out!

I don't know. If they've fucked up this early, they should probably just shut the company down.

[Darkflame]

Scepticism at this point is just weird.
Microsoft has hundreds of Hololens units at Build and developers have been getting their hands on it and build (crude) demos with Unity. The idea the "footage is a lie" at this point is just weird, and contrary to the now vaste numbers of hands on.

The downsides I listed earlier (the reduced viewing angle), have been known since January, and the other main one (controlled lighting, things too transparent) doesn't seem to be bothering people much with the hands on coming in right now.
The January hands on conforming exactly to what people are experiencing now - only with the thing in a decent looking untethered unit. (which was the main scepticism earlier, Id point out)

If the thing is "years away" at this point I couldn't for the life of me see why.

They have it in the unit they showed, they have it wireless, they have it working with people walking around.

Really not seeing any massive technical problem they are hiding here. Battery life? Who cares, its indoor only.
The rest is just scaling production, cost, and making sure theres a decent app selection.

It wont be perfect, of course not, but this thing should easily be ready by the years end. Maybe January release? Probably depends how happy Microsoft is with app's, and feedback from the various selective industry testing which apparently they already started.

The best thing is, however, it doesn't have to be perfect. The Ford Model T wasn't perfect - it was better then perfect - it was important.
Microsoft seems open for others to developer compatible devices too, so we (hopefully) will get iterative development yet with compatible apps. So if the viewing angle is a deal breaker for many, there will be pressure on the market to do better.

[Little stoned kiwi]

Microsoft’s HoloLens will cost more than your Xbox One or PlayStation 4

Which isn't surprising. But does make it years away until it comes down in price or proves itself a necessity.

[pogo]

Scepticism at this point is just weird.
Microsoft has hundreds of Hololens units at Build and developers have been getting their hands on it and build (crude) demos with Unity. The idea the "footage is a lie" at this point is just weird, and contrary to the now vaste numbers of hands on.

They were not really building apps though, just flicking a switch to enable a feature:

Microsoft had prewritten every coding script in the demo, so this meant we checked a box in Unity to turn it on

And it was not a single app but a series of apps

To demonstrate how developers will create apps for the Windows Holographic platform, Microsoft took us through a series of educational demos by loading code directly onto the HoloLens for each one of the device's key features.

The “basically a lie” refers to the footage MS showed when they first announced it. Footage is always going to be faked in a way, as unlike Oculus it doesn’t exist on a host PC or even within the device – what the user sees is a combination of rendered footage and what they actually see. But MS could have made something like the actual device, not vastly exaggerating its capabilities. That’s the lie aspect.

They need to get it in the hands of developers, so developers can code their own apps and test its limits, test how it works in real environments not the carefully prepared ones MS uses for its demos, discover what other problems it has, whether making people sick or a short battery life. They can release them with low resolution and narrow viewports: as with Oculus developers can work with that on the assumption it will be improved by release. So there must be something else holding them back from a wide or even limited release.

Only once it’s in the hands of developers does it make sense to talk about a release date, which may still be years away. Until then it’s purely a tech demo. An impressive one but with far too many unknowns to take it seriously as a development platform.

[Jamesac]

I'm sure the hololens is lovely tech and all, but I don't want it. Not in the "I'd turn down a free one" way, but in a "don't care" sense. I just don't see it as anything useful.

[gravy]

http://arstechnica.com/gadgets/2015/05/ ... f-reality/

Interesting, the same FOV complaints but no problems with overlayed graphics being too see through or sticking to the test environment.

That's promising, you'd hope the FOV could be widended before launch.

[gumboot]

Is the FOV actually bad, or is it just made to look bad because things beyond the angle of the display are still visible to remind you that there's a limit?

[pogo]

It’s odd they didn’t just do the same as Oculus, stick a lower resolution display in there so people get used to the size with the expectation that the resolution would improve, as it did in DK2 and may do again before launch. But what they are doing is rather more complex. First they need to not just display an image but display an image which matches up with the world you see. Not only the right position but the right size, the right amount of parallax. But also they aren’t displaying it directly on a screen in front of you but projecting it indirectly onto the curved surface you see through.

I suspect that last bit is hard to get right and hard to scale up. It probably needs pretty specialised optics to not only project the image but with the correct depth of field and onto a curved surface. And the bigger you make it the more complex the optics get: not only bigger but you need to do more to deal with artefacts such as spherical aberration and colour fringing. So maybe it’s been kept small not only to limit the number of pixels but to avoid optical distortion which would be very disconcerting.

[gumboot]

Certainly off-axis you can't guarantee that what you draw is aligned with the world unless you know precisely the size and shape and orientation of the eyeball. That's simple geometry.