Hacker News Reader

Comment by gnicholas | Hacker News Reader

> Given how ambitious the hardware package is, the software paradigm is surprisingly conservative.

I actually see this as a bit less surprising. After all, if you change the hardware in a big way, and you change the software in a big way, users will have a harder time adjusting to the new platform. Instead, they're making a big leap on the hardware side, keeping legacy apps and concepts, and then will presumably iterate to more 'native' experiences that were previously impossible/unimaginable/unintuitive.

View on HN

Text is still a large part of the interfaces we use.

We are all highly trained to read text, it seems basic but it is in practice quite abstract.

Text is still best read on a flat surface.

The great innovation I can see with this new Apple device is eye tracking, they have not invented it, but they might have perfected it enough to be useable.

Eyes could be better than a mouse.

Yeah, if their eye tracking plus foveated rendering works as advertised, it could be a huge step forward. I'm really curious how responsive the gesture controls will be too, it was really cool seeing the finger pinches(?) being used as an input method. I wonder if it's specifically designed just for that thing or if it's all built out to track any arbitrary hand gesture accurately. And I wonder what the language/api for describing hand gestures would even look like.

I don’t doubt them. The PSVR2 has both and it supposedly works very well.

That lacks other features of course and must be tethered to a PlayStation 5.

But eye tracking + foveated rendering seems like it’s going to become table stakes in the next few years.

Eye tracking is almost certainly more accurate and faster. To the point where we make it an entire game to see who can who can move the mouse to what their eyes are already looking at fastest, in the form of shooter games.

Interesting point — if you no longer need to use a joystick to aim your weapon, how will controllers evolve? Will the second joystick be used for some other function, or will it be replaced by a different type of input method?

It would be funny if controllers evolved to be more like the single-joystick models that we had decades ago, with the joystick on the left and rows of buttons on the right. History doesn't repeat itself, but perhaps it'll rhyme?

One joystick has to exist since there are many videogames where the character is not in focus (barely), thinking of hack and slash with top down camera (the recent diablo 4),the majority of the focus is on where the attacks are going.

In top down bullet hell shooters the player had to aim and shoot in one direction as it walks in another (potentially opposite), so one stick is still needed

Agreed — I'm thinking we'd have just one joystick, to control movement. The second joystick could evolve into some new interface, or devolve into a row of buttons.

PSVR2 already has eye tracking. And unlike the Apple Vision it also has controllers with a stick. Perhaps there are already such games which let you aim with your eyes (assuming you are shooting at something), while you move around with one stick, and the second stick isn't used at all.

In the demo, I was hoping to see some app/screen/display anchored to certain walls or being fixed in a 3D spot. That way I can walk to a place to see that "screen".

Isn’t that an intractable problem without LiDAR? I suppose the naive solution would be to fix the screens in spots as you suggest as opposed to pinning onto real life locations.

edit: ah, right the vision pro does have lidar. I look forward to seeing this pinning if it's implemented. I'm still skeptical it would work well - I've used ARKit and it's not really that accurate when you're moving around.

It should be relatively simple with a QR code or similar physical marker in the real world.

With yeah I agree. Without though it seems difficult

why do you guys say this? Amazon's "View in your room" feature does a decent job of placing a furniture in AR. We can even walk around/towards it. And that tech is just Apple ARKit right?

I don’t think it’s that good. “Decent” is accurate

right.. so maybe Apple just thinks it is not good enough yet. They must be stretched on onboard compute capacity already.

And some might argue you just switch apps to find that "screen".

Personally I would put something interesting on the ceiling and look up at it.

Is it possible to track your position that accurately in an arbitrarily large space?

But the Apple Vision has LiDAR.

It does have LiDAR, same as the iPhone/iPad AR stuff has had for years now.

The Quest has been doing this for years with the guardian setup and now with their option to bring your desk and couch into VR. I think the Pro has more advanced features but I haven’t tried it. You do have to assist it by drawing our boundaries but it’s pretty easy to do

In the demo there is a scene where a guy displays some iceberg photo on a virtual screen and then moves closer to it, with the screen anchored to some 3D point in the real world.

They kept doing that, like the woman sitting on the sofa watching a film leans down towards her popcorn[1] and the screen doesn't move, or the woman lying in bed watching the ceiling change into a blue sky[2] instead of just a blue sky where she was looking.

[1] https://youtu.be/TX9qSaGXFyg?t=180

[2] Same video, 5min 12s: https://youtu.be/TX9qSaGXFyg?t=312

I think they'll inevitably introduce this down the track. The idea that you can set up a workspace as fixed to a particular place. Sit on the couch, and by default it has media apps and a big screen. Put it on in your office, and you can cover the walls with reference drawings and appropriate apps that are all where you left them.

That was repeated a bunch of time during the keynote: "familiar"

Reply on news.ycombinator.com