This post was a bit difficult to parse; that's what it sounded like to me, although more as a interface-cleanliness feature than a rendering optimization. There also seemed to be an element of a non-euclidean/non-consistent-geometry idea there. (Which makes some sense - if you lean in or turn your head to get more UI detail you don't necessarily want it to take over your entire field of view. Although personally I think this might have the same problems as a "Minority Report"-style UI regarding long-term usability.)
I don't agree that VR should mimic the forms of reality whatsoever. I'm incredibly dysphoric about the body and would heavily prefer to express myself and not just more of what the world already sees of me. If I wanted to express more of that I'd just go outside.
>If there is a small icon on my laptop screen, no amount of me moving closer will magically add resolution. But if there is a small icon in VR, leaning in will resolve more detail.
It works for laptops too. You either are starting too close or your monitor is not high enough resolution.
>can pinch-zoom on a photo, there’s nothing you can do that lets you peer closer at an interface element and get more data than is already there
That's what the pinch zoom gesture does already.
>Like, imagine looking at the tiny wi-fi icon in the top bar on your home screen. Simply lean your head towards it a little (unconsciously) and you are able to read the name of the current wi-fi network;
The author overestimated the willingness of people to want to read small text along with not being able to sit still.
This guys blog points to his Poem "AI Clock" project.
https://poem.town/
The hardware reminds me of these:
https://lilygo.cc/products/t5-e-paper-s3-pro
https://shop.m5stack.com/products/m5papers3-esp32s3-developm...
I really struggle to picture this. Is it like LOD?
This post was a bit difficult to parse; that's what it sounded like to me, although more as a interface-cleanliness feature than a rendering optimization. There also seemed to be an element of a non-euclidean/non-consistent-geometry idea there. (Which makes some sense - if you lean in or turn your head to get more UI detail you don't necessarily want it to take over your entire field of view. Although personally I think this might have the same problems as a "Minority Report"-style UI regarding long-term usability.)
I don't agree that VR should mimic the forms of reality whatsoever. I'm incredibly dysphoric about the body and would heavily prefer to express myself and not just more of what the world already sees of me. If I wanted to express more of that I'd just go outside.
>If there is a small icon on my laptop screen, no amount of me moving closer will magically add resolution. But if there is a small icon in VR, leaning in will resolve more detail.
It works for laptops too. You either are starting too close or your monitor is not high enough resolution.
>can pinch-zoom on a photo, there’s nothing you can do that lets you peer closer at an interface element and get more data than is already there
That's what the pinch zoom gesture does already.
>Like, imagine looking at the tiny wi-fi icon in the top bar on your home screen. Simply lean your head towards it a little (unconsciously) and you are able to read the name of the current wi-fi network;
The author overestimated the willingness of people to want to read small text along with not being able to sit still.
[dead]