This article hints had why I am always suspicious of "UX" and "UI" engineers. As it notes there is a strong resistance to formalism of any sort, and working with and considering the user doesn't seem to be the first element of design rather a focus on "style" and "differentiation" seem to plague the field. Ultimately software is to be used to accomplish a task and if you aren't aiding in that are you really adding value.
Of course as I type this comment I am forced to confront my own hideous reflection as a software engineers as all of the criticisms I level at "UI" and "UX" engineering apply just as well to my own field. The difference is instead of being Bohemians about how the drop shadow effects the overall sense of user well being, we are hipsters in whatever the newest framework or language is because it introduces some old concept repackaged as a new feature and is the hot shit. Many argue that programming is more of an art than a science that a good developer can just right good code and it can really be measured, and how often is software projects, personal or professional derailed by wanting to have a "pure and elegant" architecture, or one that is hyper scale, meanwhile the business just wants crap done, and there is plenty of times that developers build something without ever talking to a user.
I don't know maybe there is some reflection to be taken out of this. On the other hand I think designers have been screwing with things to much we reached the pinnacle of UI and UX in the Unix tool chain, CLI, terminal, text based human computer interaction and everything after that is designed for wussies that aren't worthy to work a keyboard.
Can you explain why you're asking this question? It almost seems like a rherotical retort to the GP who is criticizing UI/UX designers while also introspecting their own misgivings and hypocritical aspects in software engineering. I thought it was well put.
They took button drop shadows away to punish the contemptible user.
Now they will bring them back, but with so many shadows and edges that our poor, beset-upon user, who just finally figured out where they can and can't click on a field of identical boxes called "flat ui", will experience vertigo as they try to mentally process what can and can't be done with this quasi-3d world that is presented to them.
Well, to be fair, I don’t think we ever really figured out the flat world. There is just no rhyme or reason to it; there are no affordances, so you can only learn via trial and error. Pretty much the worst system possible.
Skinny jeans are in -> Skinny jeans are out -> repeat.
Listening to designers try to justify why neumorphic is better than skeumorphic is better than oldmorphic is better than someothermorphic is like listening to teenagers try to justify why whatever jean leg style is currently "in" is fundamentally and absolutely better than whatever jean leg style their parents wore. It's just fashion, nothing more, nothing less, and certainly no more a revolution in UX than a 10% increase or decrease in the amount of fabric in the legs of a pair of jeans.
IMHO companies should include older designs, or allow people to adjust everything. So Windows 10, for instance, could come with the designs of 10, 8 and 7.
As far as I can tell, Windows 10 doesn't natively allow you to use the older window decoration styles—you'd have to purchase 3rd party software to do so.
That was a snarky dig at how Windows still has different styles and paradigms from different eras in different places. Menu bars, tool bars, control panels, and so on.
I really loathe how UI designers have become so entirely separated from the world they live in, with the result that they are forcing random change on users just for the sake of fashion, not actual innovation in how a UI works. All they talk about all day in day out is a "design language" or what "emotions" a font evokes. THIS DOES NOT MATTER!
It would be fine if there would be an opt-out (let me install my own skins like I can do on pretty much every Linux desktop, and even on older Windows versions). Let me decide how ugly or pretty I want my computing environment to be.
UI/UX designers (I'm using the term interchangeably, because hell if I know what actually separates a UX designer from a UI designer) should focus on functional improvements, not on superficial fluff like colors and fonts, or whether a button is flat or looks 3D. Stuff like this should entirely be under the control of the user.
Oh, and the Hamburger Menu is not a functional improvement.
Random changes is how you push the world forward. Otherwise you get stuck in a local maxima.
Today we're getting new buttons, some other day we'll see the rise of some new technology and will think "yeah, that's the obvious thing to do", quickly forgetting all the other ones that didn't succeed, but were created through the exact same process of random exploration.
But for UIs obviously this idea didn't work, and why should it, running in circles (e.g. changing colors, fonts, or button outlines) doesn't get you anywhere.
With the switch to a new UI paradigm (multi-touch) we lost discoverability, predictability through standard behaviour, shortcuts etc... why were UI designers not able to come up with innovations that improve the state of the art? Instead, touch UIs peaked with the first iPhone and since then became worse because more features were added without (I assume) serious usability research.
Instead of attempting to solve the actual problems that need solving, UI designers have retreated into their cozy corner of changing some completely unimportant surface details every couple of months, that's just an illusion of innovation. Smoke and mirrors.
Beyond fashion and aesthetics (which are totally valid, and what this article is mostly about), the past ten years of touch UI design have been a process of slowly eliminating the extremely obvious cues that were necessary for a population of user who had no idea how to relate to a touch screen. UI designers have been able to pack more functionality into UIs as they no longer have to beat people over the head with the idea that you can touch the controls.
As a UX designer, I've worked with plenty of things that isn't strictly UI. Unless you count functions like the support staff a kind of UI, which it arguably is (a kind of UI to the company or service offered).
At some point, I started calling myself "service designer" instead, because UX as a concept was fairly easily appropriated by visual designers (I guess what you call "UI designers") to mean "appearance."
Otherwise, what term do you use to mean that you're working to improve the experience of using a service across all interaction points it offers? I.e., across boundaries of particular screens, particular systems, particular company functions, or even particular companies in a few cases.
But I'd like to point out that it sounds like that, even while criticizing the wasteful focus on "look and feel," the look and feel seems to be exactly what bothers you, not the functionality behind it. If only functional changes mattered, we wouldn't register either way if they decided change the visuals of the system. You wouldn't need skins if the impact of it were immaterial to you.
It ends up that visuals are functional as well, and there's no hard line to draw between aesthetics and functionality. If I were to draw up a system that has great information architecture, navigation/findability, affordances, and the rest of it, but used terrible contrasts and font-weight: 100 everywhere, I would still have failed at making it usable.
I do agree that many designers seem to have some kind of pornographic relationship to UI, where they revel in it for its own sake. Personally, I consider UI to be cartilage: there should be as little as possible of it (but not less), and it should perform its function and stay out of the way as much as possible.
What "be as little as possible of it" and "staying out of the way" means is very contextual to service itself and the people using the service, which is why it is hard to establish a universally objective notion of what this means. It always ends up being an exploration to balance a vast set of variables.
Maybe Apple is completely right in making macOS more iOS-like, since familiarity is one way of staying out of the way, and they seem to want to focus on people who own an iDevice, but no Mac. But their current take on it has me worried that they've lost interest in accessibility (in all meanings of the word) as a design goal.
Google Meet - the mute/hangup buttons are hidden in the bottom bar in a completely undiscoverable way
1) Why hide them? It is like hiding the steering wheel in the car and popping it out when you need it.
2) Why not make them discoverable?
3) Why can't we give a little bit of real-estate to the most important interaction - to unmute yourself before speaking - a permanent place in the UI by sacrificing sleekness a little bit?
This is not just aesthetics, far from it. There is so much horrible UI/UX/Service/Whatever-you-wanna-call-it in the world. It is filled with it.
I could write a 10 page article about Twitter's UI/UX. It is full of obvious improvements like this. It doesn't take an expert to notice these and they're not off for a debate - there is no subtlety or subjectiveness here.
People like yourself are rare and not that common. Most UI/UX folks are all about following design trends, aesthetic sensibilities and injecting personal taste into products as if it's some kind of a creative extravaganza.
I feel the same way. Earlier this week I read a wonderful article from Bruce Tognazzini, who worked at Apple from 1978 to 1992 and was instrumental in designing the usability guidelines for the Apple II and the classic Mac OS, called "The Third User" (https://asktog.com/atc/the-third-user/), where he makes the claim that Apple's UI approach for iOS heavily focuses on buyers at the expense of users. I personally believe that this trend is commonplace throughout personal computing and on the Web these days. UI has now been tied into notions of brand identity, with some developers going as far as to demand that operating system environments respect their design choices rather than enforce system-wide conventions (see the "Please don't theme our apps" petition from some GNOME developers: https://stopthemingmy.app). There's been a shift away from the utilitarian UI designs from the 1990s (such as the classic Mac OS and Windows 95) to designs with a lot of visual appeal (such as Mac OS X's Aqua and its iterations over the past two decades, Windows Vista's Aero, Windows 10's flat design, Google's Material Design, and iOS). Unfortunately, there's a tendency for some of these visually-appealing designs to not work well for users in practice, especially for power users.
I agree with you: the solution should be to allow users to not only theme their UIs, but to even modify them as they see fit. Consider GIMP for example. While some people like it's interface, other people complain that it's un-intuitive and not like Photoshop. Well, what if GIMP's functionality was separated from its interface, and there were two modifiable files that GIMP use: one that describes the UI elements (kind of like HTML: e.g., buttons, views, menu items), and another one that describes how those UI elements are displayed (kind of like CSS). Imagine if users could modify these HTML and CSS files to create different GIMP interfaces based on their liking. Imagine if the GIMP user community shared UI configuration files amongst each other. Users will have a lot more flexibility regarding how they want to use their software tools. This will also be a tremendous boon for people with accessibility needs.
This is not just talk; I'm actually currently in the design phase of a side project where I'm building such a UI toolkit for the X Window System in Common Lisp as part of a larger goal that I have of building a OpenDoc-style (https://www.youtube.com/watch?v=oFJdjk2rq4E) component-based desktop that leverages the powerful Common Lisp Object System. The UI toolkit that I'm planning to build will define UIs using three S-expression configuration files: a content definition file, a style definition file, and a localization file.
For me it's more like uglymorphism, I really think these icons looks ugly.
I don't believe I'm saying it but I think that apple need to learn from MS with their fluent design. I understand that they trying to add shadows and dimension to flat design. I think they really distracting and look kinda like 90s.
This article defends change for the sake of change (even the style is named “neu” — new). But let’s take it seriously for a moment:
Shadows and lighting are actually cues we take into account when finding physical switches, handles, etc. yet adding them to the screen uses a lighting source “origin” that is unrelated to the real world. We literally look through a screen into an alien world.
What if our phones and computers had adequate ambient light sensing that the shadows can be Aligned to the environment? That the shadows on the screen point the same way as the shadows of the keyboard keys, or of something sitting on the desk or against the wall behind your screen?
That might actually make an interface easier and more intuitive.
> What if our phones and computers had adequate ambient light sensing that the shadows can be Aligned to the environment?
That is an interesting concept. I suspect the only times it would work would be when you are sitting near a desk lamp or are outside in the sun (directional light from a points source). Otherwise you will likely have omnidirectional lighting.
You could use the accelerometer, which would either work like those dynamic backgrounds.
Personally I'm OK with "flat" stuff in predictable places like the macos menubar - all things in a constrained space are either drop-down menus, or are for information with clickable settings
But in free-form where anything can appear, always on shadows for buttons might be boring but they are GOOD.
I'm not a fan of the current flat, 'material' design, but I do have reservations about "MacOS Dribbble". I hope the contrast issues will be fixed, and icons will be more in line with each other. In the end the UI needs to comprehensible and get out of the way of what you are trying to do.
When I saw some of the example icons my first thought was what my phone will look like if all the app/icon designers don't agree on which direction the light is coming from.
In some examples the light is coming from above you, in others it seems to come from "north" of the screen, and another from north west.
I wonder if there’s a forum full of sewing machine engineers somewhere railing against how useless fashion is and how you can’t beat a good old fashioned set of woolen monk’s robes for warmth and functionality
Aside from snarking at UX/UI designers and what is perceived as artistic pretension, there are some interesting thoughts and direction for this new style.
Assuming that AR is something that is coming to our UIs, head up displays in vehicles, AR via cameras in our phones, potentially moving to eye level, this new design language, with the concept of icons and affordances using lighting, might be an interesting approach.
When you're overlaying reality with digital additions, having them lit appropriately, taking into account the ambient lighting (not just the levels, but the directions) and using that to place the elements makes them much easier to understand.
31 comments
[ 1.7 ms ] story [ 70.0 ms ] threadOf course as I type this comment I am forced to confront my own hideous reflection as a software engineers as all of the criticisms I level at "UI" and "UX" engineering apply just as well to my own field. The difference is instead of being Bohemians about how the drop shadow effects the overall sense of user well being, we are hipsters in whatever the newest framework or language is because it introduces some old concept repackaged as a new feature and is the hot shit. Many argue that programming is more of an art than a science that a good developer can just right good code and it can really be measured, and how often is software projects, personal or professional derailed by wanting to have a "pure and elegant" architecture, or one that is hyper scale, meanwhile the business just wants crap done, and there is plenty of times that developers build something without ever talking to a user.
I don't know maybe there is some reflection to be taken out of this. On the other hand I think designers have been screwing with things to much we reached the pinnacle of UI and UX in the Unix tool chain, CLI, terminal, text based human computer interaction and everything after that is designed for wussies that aren't worthy to work a keyboard.
Now they will bring them back, but with so many shadows and edges that our poor, beset-upon user, who just finally figured out where they can and can't click on a field of identical boxes called "flat ui", will experience vertigo as they try to mentally process what can and can't be done with this quasi-3d world that is presented to them.
The design community always manages to surprise.
Listening to designers try to justify why neumorphic is better than skeumorphic is better than oldmorphic is better than someothermorphic is like listening to teenagers try to justify why whatever jean leg style is currently "in" is fundamentally and absolutely better than whatever jean leg style their parents wore. It's just fashion, nothing more, nothing less, and certainly no more a revolution in UX than a 10% increase or decrease in the amount of fabric in the legs of a pair of jeans.
It would be fine if there would be an opt-out (let me install my own skins like I can do on pretty much every Linux desktop, and even on older Windows versions). Let me decide how ugly or pretty I want my computing environment to be.
UI/UX designers (I'm using the term interchangeably, because hell if I know what actually separates a UX designer from a UI designer) should focus on functional improvements, not on superficial fluff like colors and fonts, or whether a button is flat or looks 3D. Stuff like this should entirely be under the control of the user.
Oh, and the Hamburger Menu is not a functional improvement.
Today we're getting new buttons, some other day we'll see the rise of some new technology and will think "yeah, that's the obvious thing to do", quickly forgetting all the other ones that didn't succeed, but were created through the exact same process of random exploration.
With the switch to a new UI paradigm (multi-touch) we lost discoverability, predictability through standard behaviour, shortcuts etc... why were UI designers not able to come up with innovations that improve the state of the art? Instead, touch UIs peaked with the first iPhone and since then became worse because more features were added without (I assume) serious usability research.
Instead of attempting to solve the actual problems that need solving, UI designers have retreated into their cozy corner of changing some completely unimportant surface details every couple of months, that's just an illusion of innovation. Smoke and mirrors.
At some point, I started calling myself "service designer" instead, because UX as a concept was fairly easily appropriated by visual designers (I guess what you call "UI designers") to mean "appearance."
Otherwise, what term do you use to mean that you're working to improve the experience of using a service across all interaction points it offers? I.e., across boundaries of particular screens, particular systems, particular company functions, or even particular companies in a few cases.
But I'd like to point out that it sounds like that, even while criticizing the wasteful focus on "look and feel," the look and feel seems to be exactly what bothers you, not the functionality behind it. If only functional changes mattered, we wouldn't register either way if they decided change the visuals of the system. You wouldn't need skins if the impact of it were immaterial to you.
It ends up that visuals are functional as well, and there's no hard line to draw between aesthetics and functionality. If I were to draw up a system that has great information architecture, navigation/findability, affordances, and the rest of it, but used terrible contrasts and font-weight: 100 everywhere, I would still have failed at making it usable.
I do agree that many designers seem to have some kind of pornographic relationship to UI, where they revel in it for its own sake. Personally, I consider UI to be cartilage: there should be as little as possible of it (but not less), and it should perform its function and stay out of the way as much as possible.
What "be as little as possible of it" and "staying out of the way" means is very contextual to service itself and the people using the service, which is why it is hard to establish a universally objective notion of what this means. It always ends up being an exploration to balance a vast set of variables.
Maybe Apple is completely right in making macOS more iOS-like, since familiarity is one way of staying out of the way, and they seem to want to focus on people who own an iDevice, but no Mac. But their current take on it has me worried that they've lost interest in accessibility (in all meanings of the word) as a design goal.
1) Why hide them? It is like hiding the steering wheel in the car and popping it out when you need it.
2) Why not make them discoverable?
3) Why can't we give a little bit of real-estate to the most important interaction - to unmute yourself before speaking - a permanent place in the UI by sacrificing sleekness a little bit?
This is not just aesthetics, far from it. There is so much horrible UI/UX/Service/Whatever-you-wanna-call-it in the world. It is filled with it.
I could write a 10 page article about Twitter's UI/UX. It is full of obvious improvements like this. It doesn't take an expert to notice these and they're not off for a debate - there is no subtlety or subjectiveness here.
People like yourself are rare and not that common. Most UI/UX folks are all about following design trends, aesthetic sensibilities and injecting personal taste into products as if it's some kind of a creative extravaganza.
I agree with you: the solution should be to allow users to not only theme their UIs, but to even modify them as they see fit. Consider GIMP for example. While some people like it's interface, other people complain that it's un-intuitive and not like Photoshop. Well, what if GIMP's functionality was separated from its interface, and there were two modifiable files that GIMP use: one that describes the UI elements (kind of like HTML: e.g., buttons, views, menu items), and another one that describes how those UI elements are displayed (kind of like CSS). Imagine if users could modify these HTML and CSS files to create different GIMP interfaces based on their liking. Imagine if the GIMP user community shared UI configuration files amongst each other. Users will have a lot more flexibility regarding how they want to use their software tools. This will also be a tremendous boon for people with accessibility needs.
This is not just talk; I'm actually currently in the design phase of a side project where I'm building such a UI toolkit for the X Window System in Common Lisp as part of a larger goal that I have of building a OpenDoc-style (https://www.youtube.com/watch?v=oFJdjk2rq4E) component-based desktop that leverages the powerful Common Lisp Object System. The UI toolkit that I'm planning to build will define UIs using three S-expression configuration files: a content definition file, a style definition file, and a localization file.
Shadows and lighting are actually cues we take into account when finding physical switches, handles, etc. yet adding them to the screen uses a lighting source “origin” that is unrelated to the real world. We literally look through a screen into an alien world.
What if our phones and computers had adequate ambient light sensing that the shadows can be Aligned to the environment? That the shadows on the screen point the same way as the shadows of the keyboard keys, or of something sitting on the desk or against the wall behind your screen?
That might actually make an interface easier and more intuitive.
That is an interesting concept. I suspect the only times it would work would be when you are sitting near a desk lamp or are outside in the sun (directional light from a points source). Otherwise you will likely have omnidirectional lighting.
You could use the accelerometer, which would either work like those dynamic backgrounds.
Personally I'm OK with "flat" stuff in predictable places like the macos menubar - all things in a constrained space are either drop-down menus, or are for information with clickable settings
But in free-form where anything can appear, always on shadows for buttons might be boring but they are GOOD.
I'm not a fan of the current flat, 'material' design, but I do have reservations about "MacOS Dribbble". I hope the contrast issues will be fixed, and icons will be more in line with each other. In the end the UI needs to comprehensible and get out of the way of what you are trying to do.
That'd be cool.
/s (in case it's not obvious)
In some examples the light is coming from above you, in others it seems to come from "north" of the screen, and another from north west.
Assuming that AR is something that is coming to our UIs, head up displays in vehicles, AR via cameras in our phones, potentially moving to eye level, this new design language, with the concept of icons and affordances using lighting, might be an interesting approach.
When you're overlaying reality with digital additions, having them lit appropriately, taking into account the ambient lighting (not just the levels, but the directions) and using that to place the elements makes them much easier to understand.