4 ms·

The nice thing about the "linear" spaces on MacOS and Windows is that in the vast majority of cases there's a 1:1 mapping between the pixels on the screen and t

The nice thing about the "linear" spaces on MacOS and Windows is that in the vast majority of cases there's a 1:1 mapping between the pixels on the screen and the UI elements. What makes me nervous is when they start making the pointer space non-linear. Is there a risk that you just can't get the cursor to activate a part of the screen because it keeps getting sucked into a nearby control? Or is the pointer space just a reversibly-warped version of the screen space?

ricardobeat 7y ago
Most likely the Human Interface Guidelines that Apple enforces for all of their software is a major enabler for this, as it mandates minimum touch areas and spacing that should eliminate that type of issues. No other platform has as much native UI adoption - used to be the case for Mac too until Electron and others came along.
- Nition 7y ago
  Do minor platforms count? I suspect Sailfish has higher native UI adoption. IIRC back when I had a Sailfish phone, all the apps I used including the third-party ones used the same standard UI elements.
  - thomashobohm 7y ago
    They probably meant out of the platforms Apple competes with. Presumably I could find a platform with, say, 12 users and only 1 non-native app, and if that app followed native UI guidelines I could say it has 100% native UI adoption and thus the “highest” out of every platform ever, but that is a useless statement.
- saagarjha 7y ago
  You forgot watchOS, where you literally cannot design a non-native app.
skavi 7y ago
I’d expect them to hand more precise control back to the user as the pointer slows down.
lonelappde 7y ago
But there isn't a 1:1 mapping between where I am trying to tap and where my fat finger registers a click on a linear screen.
kondro 7y ago
I have one of the new Magic Keyboards in front of me now. I had the same reservations as you before trying it out (and it’s a little hard to tell because all I have to rely on is my meat brain’s interpretation), but it feels like the cursor still moves in linear space. It’s just that whilst it’s over a button (or other control) that the cursor disappears and the selection feels glued to the control for the duration it passes over.
bitwize 7y ago
There hasn't been a linear mapping between on-screen pixels and space traversed by the mouse for decades now. "Mouse acceleration" messes that up.
- userbinator 7y ago
  Mouse acceleration is one of the things I always turn off on a clean install, precisely because of that nonlinearity. When I have to use others' systems who do have it on, I find myself always overshooting with the cursor.
systoll 7y ago
The 'pointer magnetism' in iPadOS has two different effects. The main effect doesn't change pointer movement: Most controls on iPadOS are bigger than they look, because fingers. With no adaptation, the cursor would be hovering over controls when it looks like it's 24px away from them. To avoid this, controls move toward the cursor when hovered, depending on how far away the cursor is from their centre. On any given frame, if the cursor is within a control's actual bounds, the control will visually overlap the cursor's current position. The second thing does affect pointer movement: If you quickly swipe in a single direction, and the cursor lands on the edge of a control, effect #1 still happens on the approach, but the element bounces back to its resting location, and the cursor goes with it. This seems to work well in practice. It doesn't change anything while your finger is moving, it doesn't affect any 'precise' motions, and it can't lead to your swipe landing on a different element.

ricardobeat 7y ago

Most likely the Human Interface Guidelines that Apple enforces for all of their software is a major enabler for this, as it mandates minimum touch areas and spacing that should eliminate that type of issues. No other platform has as much native UI adoption - used to be the case for Mac too until Electron and others came along.

Nition 7y ago
Do minor platforms count? I suspect Sailfish has higher native UI adoption. IIRC back when I had a Sailfish phone, all the apps I used including the third-party ones used the same standard UI elements.
- thomashobohm 7y ago
  They probably meant out of the platforms Apple competes with. Presumably I could find a platform with, say, 12 users and only 1 non-native app, and if that app followed native UI guidelines I could say it has 100% native UI adoption and thus the “highest” out of every platform ever, but that is a useless statement.
saagarjha 7y ago
You forgot watchOS, where you literally cannot design a non-native app.

skavi 7y ago

I’d expect them to hand more precise control back to the user as the pointer slows down.

lonelappde 7y ago

But there isn't a 1:1 mapping between where I am trying to tap and where my fat finger registers a click on a linear screen.

kondro 7y ago

I have one of the new Magic Keyboards in front of me now. I had the same reservations as you before trying it out (and it’s a little hard to tell because all I have to rely on is my meat brain’s interpretation), but it feels like the cursor still moves in linear space. It’s just that whilst it’s over a button (or other control) that the cursor disappears and the selection feels glued to the control for the duration it passes over.

bitwize 7y ago

There hasn't been a linear mapping between on-screen pixels and space traversed by the mouse for decades now. "Mouse acceleration" messes that up.

userbinator 7y ago
Mouse acceleration is one of the things I always turn off on a clean install, precisely because of that nonlinearity. When I have to use others' systems who do have it on, I find myself always overshooting with the cursor.

systoll 7y ago

The 'pointer magnetism' in iPadOS has two different effects. The main effect doesn't change pointer movement: Most controls on iPadOS are bigger than they look, because fingers. With no adaptation, the cursor would be hovering over controls when it looks like it's 24px away from them. To avoid this, controls move toward the cursor when hovered, depending on how far away the cursor is from their centre. On any given frame, if the cursor is within a control's actual bounds, the control will visually overlap the cursor's current position. The second thing does affect pointer movement: If you quickly swipe in a single direction, and the cursor lands on the edge of a control, effect #1 still happens on the approach, but the element bounces back to its resting location, and the cursor goes with it. This seems to work well in practice. It doesn't change anything while your finger is moving, it doesn't affect any 'precise' motions, and it can't lead to your swipe landing on a different element.