Live data from Hacker News

Notes on Vision Pro

notes.andymatuschak.org

271–280 of 700 posts

Re: Notes on Vision Pro

#271

His comments on anchoring virtual tools in physical space reminds me of the work Microsoft was doing with Hololens and Windows Mixed Reality. If I recall correctly, there were APIs for anchoring things in space.

I recall stumbling upon those too in the Win 10 mixed reality API docs. There were a lot of interesting things in there that never really saw the light of day/any significant use

EDIT: Found the API docs. There's some really interesting classes in there! I remember handling coordinates to be confusing/a pain though haha. https://learn.microsoft.com/en-us/uwp/api/windows.perception...

Re: Notes on Vision Pro

#272
I find it interesting that Microsoft had _so much of this_ 9 years go. I developed with the HoloLens 1 around 2016 and I recall:

* Fingers together gesture clicking.

* Voice activated menu navigation. It was glitchy/I never used it though. In 2023 these sorts of systems are much better though.

* No controllers. It was all gesture based. Opening the start menu required your hand upturned, fingers together then outstretch. Kind of like an "open" gesture.

* "Pointing" based on head looking vector, which was annoying.

* Spatial anchors and being able to remember past spaces and how you used them. There was a whole set of SDK APIs for the spatial stuff built into Win 10

The Vision Pro iterates on some of these. Eye tracking for pointing and more cameras for tracking hand pose for clicking, to reduce the annoyance/strain. IMO all mandatory for long term usage (I frequently got motion sick with the HoloLens and the 3 months of debugging/developing on it were a challenge).

Maybe if Microsoft had switched to more commonplace display technology and continued to iterate on their product instead of let it languish, they could have had a solid competitor to this, if not maybe be first to the market.

Re: Notes on Vision Pro

#273
post #243

Earlier quoted context omitted.

I imagine there will be some developers that try to have analytics fire when certain events happen, correlated with eye-hovering. Hopefully Apple will catch these tricks before they can go live. The potential value of knowing where someone's eyes are (whether they saw an advertisement, for example) could be very high, so I envision a cat-and-mouse game.

Apple addressed that in the keynote. There’ll be an invisible canvas over the window content, and the app never learns where you look (nor Apple, ie it is not further stored or processed). Only when you click (ie tap your thumb and index together) does the app get an event. (Downside is that one loses the capability for hovering, but strikes me as the correct trade off.)

Sure, that works for certain contexts. But what about a game, where your vision is used to aim a weapon? The app has to know where you are looking in order to know where to shoot. If it's an online game, then the information cannot be isolated on the device since you're shooting at other players.

Now imagine the game has advertisements, or loot boxes, or something else that the game company wants to know if you've looked at. How does Apple prevent them from learning that, if they're able to know what they need to know for the game to function as a regular FPS?

Re: Notes on Vision Pro

#274

Earlier quoted context omitted.

The interaction paradigm you describe (where it's literally impossible to look at one window while actively using another) makes so little sense that we can almost certainly rule it out. Certainly no press members who demo'd the unit are describing the interface that way. That's not "defending Apple", that's just... gosh, I can't think of any word besides common sense. From all of the presentations and press accounts…

I don't disagree with your other points AT ALL, but given how repeatedly proven that common sense is entirely a fiction, if any argument falls back on it you have to be immediately suspect of it.

I hear that now and then, but looking at the wiki on it, the “criticism” paragraph is only like 3 lines with no real refutation, out of a hefty lengthy article.

What’s your motivation for saying that? Personally I am suspect of people say there isn’t common sense.

Here is the definition at the top: Common sense (or simply sense) is sound, practical judgment concerning everyday matters, or a basic ability to perceive, understand, and judge in a manner that is shared by (i.e., "common to") nearly all people.

One can certainly say there is common sense. For instance no one likes pain, people like pleasure. People care about entertainment, not boredom. People tend to learn on the whole as they grow up. I also think on the whole have a perception of what is wrong and right, even if there is variance between cultures. The golden rule of reciprocity seems to be near universal (like: don’t kill, don’t steal and so on)

Lots of things we share as a species.

Re: Notes on Vision Pro

#275
post #186

Earlier quoted context omitted.

>Why does everyone have MacBook Pros these days? Prior to Apple Silicon, it certainly wasn't the hardware. Oh, I don't know about that. Admittedly there were some awful choices made between 2015 and the emergence of M1 architecture, but Apple's reputation for making well engineered hardware goes at least as far back as to when they started milling Macs out of aluminum. I'm thinking back specifically to the PowerMac G…

I haven't liked the usability of the entire Macbook line though, and It's very upsetting that so many laptop makers have decided to just copy it. There's lots of little details which are well made, but bad decisions: the screen hinges tend to be a little too floppy, the trackpad is oversized, removing the physical buttons makes it hard to use, moving the power button from a separate area to the "eject" button locatio…

Two thoughts here. The first is that this comments is totally valid and mostly the inverse of everything I think, really shows that we all value different things so thank god for a free market.

The two things I agree on for the 2016-2022 laptops are the sharp edges and terrible keyboards. Maybe take a look at the new laptops next time you get a chance, the edges on my MBAir and 16MBPro from 2022 are waaaay more comfortable than my previous two MacBooks and the keyboard feels much nicer too.

Back to point one though, to me the trackpads are well sized, the hinges are smooth and just right and the power button being where it is is a none issue, I even quite liked the touchbar.

Re: Notes on Vision Pro

#276
post #50

This is a bit of a rant, but I bought a Quest Pro last year, and I feel it was a total waste of money. It's actually embarrassing how much of an unpolished piece of garbage it feels like. Sure, it's fun as a PCVR headset, but as a productivity tool or as a multimedia consuming device, or as a social tool, it completely misses the mark. And I'd say that 80% of its failure can be attributed to bad software. It's buggy,…

This is on brand for meta. Look at fb. It didn't come with instructions, but that didn't stop Cambridge Analytica from happening. Look at React. It didn't have instructions and yet it blew up. I bet Meta expected people to pick up the slack, that the best of what was there while they killed time polishing it.

Zuck was probably part of the problem... a room full of POs pushing for stupid shit while building on a rocky foundation.

Re: Notes on Vision Pro

#277
post #270

But it does put an enormous amount of pressure on the eye tracking. As far as I can tell so far, the role of precise 2D control has been shifted to the eyes. I've got one good eye and one bad eye. The bad eye is legally blind, has an off-center iris and is kind of lazy w.r.t. tracking. I'm extremely curious to know how Vision Pro deals with this. One certainly hopes there's some kind of "single eye" mode; certainly s…

> The Minority Report movie probably set the industry back by a decade or two. Waving your hands around to control stuff seems logical but is quickly exhausting. The bad idea had waves even before that: Remember the gorilla arm. http://www.catb.org/jargon/html/G/gorilla-arm.html

I remember the gorilla arm but I found out that occasionally raising a hand from the keyboard and tapping the screen is not a noticeable effort and it's faster than aiming at the touchpad or (of course) reaching for the mouse. I discovered it mostly by using a Samsung tablet with both a Dex interface over Android and with an Ubuntu Unity one (the defunct Dex for Linux.)

That's maybe because the screen is small and low. Raising an arm at or above the head is a different matter. If you're cycling with a heart rate monitor you discover that you add some beats when you raise your hands (say, go from 150 to 155) and gain some if you lower your chest close to the handlebar. This doesn't take into account increased or reduced drag because I measured it at home on trainers.

Re: Notes on Vision Pro

#278

Earlier quoted context omitted.

In the iOS vs Android case, iOS does in fact typically bring at least 50% more profit to developers, often far more.

It's more complex than that...the whole limitation on promoting your existing membership services, the rules enforced on pricing and passing the 30% to the users etc. makes it a much muddier picture. iOS sure brings some value to the table. How much actually ? who knows... For comparison Youtube is also a platform regularily criticized by its participating creators, but you're not seeing high profile lawsuits or the…

No it’s not more complicated. iOS brings in twice as much app revenue than Android in total.

What you’ve said doesn’t change this at all.

Re: Notes on Vision Pro

#279
post #270

Earlier quoted context omitted.

> The Minority Report movie probably set the industry back by a decade or two. Waving your hands around to control stuff seems logical but is quickly exhausting. The bad idea had waves even before that: Remember the gorilla arm. http://www.catb.org/jargon/html/G/gorilla-arm.html

I remember the gorilla arm but I found out that occasionally raising a hand from the keyboard and tapping the screen is not a noticeable effort and it's faster than aiming at the touchpad or (of course) reaching for the mouse. I discovered it mostly by using a Samsung tablet with both a Dex interface over Android and with an Ubuntu Unity one (the defunct Dex for Linux.) That's maybe because the screen is small and lo…

I've noticed the same thing when I ride in my basement. If I come up off of the hoods, grab a water bottle and take a drink my heart rate bumps up ~5bpm. It'll climb a bit more if I eat something like a granola bar or a gel.

What's interesting is that unlike steady state to a hard sprint, the heart rate rise is almost immediate. There's some delay - 20 to 45 seconds - if I change it up and do a sprint interval before my heart rate reflects that effort.

Re: Notes on Vision Pro

#280
post #258

Earlier quoted context omitted.

I think it was MKBHD who seems to know his stuff

Yeah. Also, the "why don't they just" (use plastic/carbon) came from me, not him. Apple is long known for using metal in places where it unnecessarily increases weight, because it makes a product feel more premium.

He did mention that competitors use plastic precisely because it is lightweight.
Post reply on HN