Just thought? No, "we've done tons of user testing" and reached the conclusion "it doesn't work". Watch the full video, not even a minute long.

https://www.youtube.com/watch?v=7zfir0Ide0A

Same conclusion that everyone has reached over the past four decades, at least. It even got a nickname.

https://en.wikipedia.org/wiki/Touchscreen#%22Gorilla_arm%22

Current Apple should show their new tests that demostrate the opposite.

(Edit) The full transcription of the video, faster than watching:

"You know, the first thing you think about is this, right? We thought about this years ago. We've done tons of user testing on this and it turns out it doesn't work. Touch surfaces don't want to be vertical. It gives great demo. But after a short period of time, you start to fatigue and after an extended period of time, your arm wants to fall off. It doesn't work. It's ergonomically terrible. Touch surfaces want to be horizontal. Hence, pads. For a notebook, that's why we've perfected our multi-touch trackpads over the years." -- Steve Jobs, 2010

I don't think vertical vs horizontal is quite the thing - people don't always hold their phones and tablets horizontally while using them - often they'd be almost vertical. But standard laptop vertical screen position is uncomfortable to use as a touch screen. I'm curious how that research would relate to how people normally use handheld touchscreens.

Yet Qubes OS demostrates it's possible to run X11 programs isolated from each other. They had the need and will to improve security, and found the way over a decade ago.

Making server side decorations an important part of it. Oh, the irony.

Qubes OS achieves isolation by separating apps into VMs. Wayland does this at the display protocol level. If Qubes OS is serious about security, they should consider Wayland with containerization. It would result in a similar level of isolation with significantly lower resource overhead.

I feel like the sheer volume of OS LPEs and escape CVEs in container runtimes indicates that, at least for now, the security boundary capabilities of containers are inferior to those of VMs.

Which is ironic, given that a lot of the tools that underly a container runtime were originally designed to facilitate security, not ease of deployment.

You're aware of the VM escape issues over the past decade? There's no perfect, and high overhead from running many guests can also create security risks. Wayland with all of its problems remains a logical step forward from the X model.

I was narrowly responding to the point about Qubes using VMs. I have no problems with Wayland’s design; I agree that it’s a step forward.

Ah, got it! Thanks for clarifying. That makes a lot of sense.