As a counter-argument, HCI was investigated pretty thoroughly in the 80s and 90s, and operating systems of the time actually had the results of that well implemented in them. I feel that modern OS developers seem determined to throw away all these lessons.
Don't get me wrong, I think the modern HCI on mobile phones is remarkably good. But I haven't seen any improvement (except maybe the mouse scroll wheel and having a higher resolution screen) on real computer interfaces since the 90s.
And then you have some real useful psychological theories on attention and user-guiding that are used for evil to create antipatterns. I don't think we're making progress.
I think we should be careful to distinguish the question of if we are growing knowledge and the question of if we are using the knowledge (and if we are using it positively). If we aren't using it, there is an interesting question of why, but I think there should be a clear difference between not finding knowledge and not utilizing the knowledge we find.
I've a theory that most UX/UI developers started in their youth as gamers, especially in "twitch" genres, because many interactions for me are now closer to playing Descent than typing a paper into Wordperfect.
> Don't get me wrong, I think the modern HCI on mobile phones is remarkably good.
One of the challenges of psychology is individual variation. Humans have more in common with one another than we have differences, but individuality is a major factor that forces psychologists to look at things statistically unless they are specifically trying to understand or control for individual variance.
I bring this up because my personal subjective opinion is that HCI on modern mobile phones is absolutely atrocious and I don't use a smart phone as much as most people as a result.
I think that when it comes interacting with a tool, what you are accustomed to makes a huge world of difference. I grew up with Desktop computers and laptops. With keyboards, in other words. As a coder and a *nix "power user", I like command line interfaces. I like being able to tweak and customize and configure things to my liking. When I have to use Macbooks at work, it has been soul crushing to me while for others they absolutely love the UI of MacOS.
I also remember the shift of the mobile revolution. A lot of us at the time were starting to get very annoyed by the creep of mobile design conventions making their way into non-mobile contexts. At the time it was understood that those mobile design decisions were "forced" as a result of the limitations of a mobile device, and it was clear that applying them to non-mobile contexts was a cost-cutting measure (mobile first, in other words).
Although well designed iconography can transcend language barriers and facilitate communication, I find that the limited resolution of a smart phone screen forcing designers to use glyphs instead of written text is very confusing to me. I mean, don't get me wrong, I would love to learn ancient Egyptian, but it is often far from intuitive or obvious what these hieroglyphs on the screen are meant to communicate to me. In other words, the iconography is not well designed IMO. At least not in a way that creates an intuitive experience FOR ME.
But a kid who grew up in a world of smart phones is going to be able to navigate them intuitively because they have years of learning what those esoteric glyphs on the touch screen are. They've had years of "typing" out text messages on a tiny touch screens.
On a good mechanical keyboard I can type upwards of 117wpm before I start making mistakes. When trying to text my wife one sentence I need to put aside an afternoon out of my day to get it written correctly. I could get started on how awful auto-correct is but everyone knows this to the point where it's become a cultural meme. Sorry, auto-correct turned "Can you grab me some milk while you're there?" into "fyi the police are here with a search warrant."
So yeah, big tangent off of "HCI on mobile phones is remarkably good." Maybe it is in a relative sense and is as good as it can get... I mean we've had years to iterate and make improvements. But I suspect that a lot of it has to do with people just learning and getting used to haphazard design decisions that just became the defacto for mobile because the tech industry (and business at large if we're being honest) loves to copy.
I also was raised on using a keyboard to interact with a computer. I agree with a lot of your points - the UI on a mobile phone is not very good at doing text-based stuff, but I think that's OK, because I shouldn't be trying to do large scale text-based stuff on such a tiny screen with a tiny input area.
What works well on the phone UI is the way that the touchscreen has been integrated well into it, and the various gestures are mostly highly intuitive as to what they do (although if we could stop maps rotating when we try to zoom in/out with the pinch gesture, that'd be lovely thanks).
The problem comes when trying to apply the mobile phone style UI to a real computer with a keyboard, large screen, etc. That's just awful, but it appears to be the route that UI designers are galloping down these days.
Interesting. I agree with you about being most comfortable with a desktop/keyboard interface, but have the exact opposite opinion regarding macs.
IMO, OSX is the perfect platform for a keyboard-driven power user. It's unix/BSD based, so software works mostly the way you want it to, but unlike Linux it "just works" without endless fiddling. I don't use the OS UI much at all: Spotlight lets me open any app with a few keystrokes. All my time is spent in the terminal or the browser.
I've done almost no fiddling on NixOS in the last 7 years. People fiddle on Linux because they like to fiddle. My experience is it absolutely Just Works. By contrast, I've had OSX at work delete my data after one update and corrupt its install after another.
Don't get me wrong, I think the modern HCI on mobile phones is remarkably good. But I haven't seen any improvement (except maybe the mouse scroll wheel and having a higher resolution screen) on real computer interfaces since the 90s.
And then you have some real useful psychological theories on attention and user-guiding that are used for evil to create antipatterns. I don't think we're making progress.