Who cares about user interfaces? (by alaric)
The Smartphone
Smartphones were a paradigm shift. A portable computer, with you 24/7, unlike even a laptop that you had to get out of your bag and open up? A small screen, with a tiny or no keyboard? Touch or a wheel, rather than a mouse?
Clearly, you couldn't just port GNOME or Windows 2000 or Mac OS X to the device and expect it to work; a small screen and different input hardware meant that existing software - written with keyboards, mice, and large screens in mind, because the GUI toolkits exposed too much of the specifics of the hardware to the application - wouldn't work very well. New user interfaces had to be developed, and so they were. Based around touch screens, they used large icons, and gestures such as swiping and pinching. Apple's iOS went through that skeumorphic phase, but I'm not so interested in visual design as the underlying conceptual model. And the model of smartphone apps was distinctly cruder than of desktop apps.
Now, the conceptual user model of a desktop UI, although quite variable between implementations, generally involved:
Some concept of "place", where the user's files lived. This could start with places like the desktop, or a My Documents of some kind. Some kind of file manager / explorer / finder interface would be provided to find and manage your files.
Some concept of "apps". When you found a file in the "place", there would usually be a default application that owned the file based on its type, which would be the one that would provide the file's icon, be the app that would open the file if you just double-clicked the file, and so on. But often you could manually stick a file into other applications that were also able to process it. Also, you could usually start apps directly - rather than via opening a file they "own" - to either create new files from scratch, or to use non-file-based functionality of the app, like playing a game or browsing the Web or something.
Maybe some notion of "screen furniture", such as a Start menu, taskbar, etc; often used as a place to invoke apps from (although you can often find them in a "place"), and to include other high-level system control functions (shut down, change settings), and to put useful information displays such as clocks.
Due to historical baggage going back to pre-graphical computing, running apps had full reign of everything in your computer (or your user account, in more locked-down systems), so any app could read or write any file if it wanted to.
Smartphones, at least to begin with, didn't really need the "file" concept, so the user interface model they adopted was much more just about "apps". The home screen would be a list of icons, representing apps, and maybe also with some furniture as well (after all, we need to show the battery level, time, wireless connection status, etc) somewhere. Each app had its own internal storage, rather than having access to a shared filesystem full of files, and with a few exceptions they couldn't access other app's data.
Yes, I know that Android and, I believe, iOS have a full UNIXy filesystem under the hood, but apps access to it is restricted and only nerds know it's there. Also, I know there are some exceptions to that rule; shared storage areas for media galleries and downloaded files exist, with restricted interfaces to them. And I suppose I should mention that your contact list is, also, available as a shared resource between applications.
This was interesting, because it was at least novel, exploring new ground. It also offered significant security advantages, catching up with the realities of downloading software from the Internet. It wasn't necessarily a terrible idea to run a random mobile app you'd downloaded, as it couldn't steal and then trash everything in your device; it would only have access to its own data storage unless it explicitly asked for permission to touch your shared areas, such as the media gallery or your contact list. This was essential to enable the mobile app ecosystem to develop.
But it also represented a shift in how computers were used, and one that various people were making money from, so it was quietly pushed and encouraged: these were much more passive consumption devices. The only high-bandwidth input device on a smartphone was the camera, so people shared photos and videos; they didn't write documents. The phone made it easy to watch videos and doomscroll. The fact that the apps had a very simple model for managing your "stuff" on the phone, with only a limited sense of "place", was fine because the only real things you wanted "on" your phone were your contacts and your media gallery (ok, a music gallery was popular for a while - then things moved more towards streaming). This, in turn, made smartphones not ideally suited for creative tasks (other than photography and videoing), or knowledge management.
However, on top of that conceptual model, there was a brief spurt of innovation in fundamental user interfacing actions. Scroll wheels were explored as a menu navigation device, and I'll include things like Graffiti text entry although it was strictly from the PDA era that predated the smartphone, because I think smartphones built on top of PDA notions and brought them into a broader mainstream, so I count them as part of the same spiritual family. All the touch-based scrolling and pinching stuff evolved out of this era, as well as new on-screen conventions to guide the user as to what actions were available.
It was fun for a while, but it stabilised pretty soon, and then progress dried up.
Tue 8th Jul 2025 7:39 pm 
Subscribe