In 90’s, skeuomorphic trend was picking up steam reaching a climax in mid-2000’s. Web 2.0 design elements were springing up, taking advantage of css; while UI elements sought transparency and blur (Windows Vista). That trend passed and then we had the whole Material design boom in 2012-2015, the entire web was sabotaged by Bootstrap. Color themes and palettes were very popular. Cookie-cut banners, slideshows, carousels, headers, etc. Post bootstrap world saw some radical shift in small circles: monospaced fonts, brutalism, vintage homage and low-res shenanigans whilst the mainstream design kept on drinking the material koolaid up until 2016.
Boom comes in Stripe with sleek blue-purple gradients, smooth animations, Bumla-lizing the entire fucking internet. It is 2019 and every new Saas service looks like Stripe, following trends like a mindless herd. The rest unfortunate (or unskilled to match Stripe standards) are still catching up with single page scrolling websites with full-width banners, stupid jumbotrons and loud obnoxious typography.
Traditional graphic design has largely been “solved” by the 1960’s - aka, the International style or the Swiss style for majority of needs. Of course you’re not gonna have Univers on a grunge band poster. For the curious, google some posters by Josef Muller-Brockmann from the 50’s - they still appear to be designed in 2019. That’s timelessness (or cynically put, blandness). He exemplified great understanding of contrast, scale and hierarchy in design to convey information. I recently watched Dieter Rams’ documentary and he is just completely disappointed by majority of design in the world. I am not a designer but I feel the same way. Rams’ design will never be “outdated” because his principles of design weren’t based on trends.
Let’s go Dark Themes! Also, next year is the death of jumbotrons. Mark my words.
Traditional development has largely been “solved” by the 1960’s - aka, BASIC or C for majority of needs. Of course you’re not gonna have FORTRAN on a modern website. For the curious, google some books by Brian W Kernighan - they still appear to be relevant in 2019. That’s timelessness.
In all seriousness, it's a mistake to say that 1 style -- The International/Swiss Style -- can solve all of design's needs. You are purposely forgetting that design relies heavily on taste. Not the designer's taste, but the customer's taste. When I go shopping for organic shampoo, I have an idea of what "organic shampoo packaging" looks like. It's beige, probably has a plant on it, etc. Tastes change, and they change constantly and organically. Changing tastes are a moving target for designers to hit, or if they're lucky they can arrive there early. Also, it's a sign of "freshness" to update your design from time to time. People expect things to update and "get better", both visually and functionally. Similar to fashion, design does not always have to be rationale and objectively beautiful -- it just has to suit it's customer's tastes.
It is a mistake to equate software development with graphic design. To compare C with International Style sounds promising on the surface, rings all kinds of bells but your analogy breaks down as follows.
International Style is not a lower abstraction or “building” block for design. It is a framework of principles.
Not only that, it ignores how culturally-dependent design is. Sometimes in really big, fundamental ways -- take, for example, the fact that in China, the color red is heavily associated with luck, joy, and happiness. A big part of design is understanding and incorporating these kinds of local knowledge; at the end of the day, design is a process, not a destination.
Much like art, it's impossible to truly separate design from the greater context of the world at large.
Dieter Rams designs were absolutely based on trends, and trends which very much fell or of favour for twenty or thirty years. The fact that they are on trend again does not invalidate that. Nor does the fact that his designs in particular are remembered and copied because he was one of the best of the era.
He was one of the best because he used an approach which was exploratory, multi-disciplined and human focused. (Tellingly, he was educated in interior design and architecture.) To suggest that graphic design is solved, is like saying a radio is solved. Technological changes coincide with changes in human understanding, taste, response and feedback.
And then to use Dieter Rams as an example, because presumably you like his aesthetic, is to not understand design at all. Or certainly not his practice of it.
On the other hand, i totally think someone could start a Rams framework (if it hasn't already been done). An unholy mixture of skeumorphic dials and material backgrounds. It could be a beautiful abomination.
"He was one of the best because he used an approach which was exploratory, multi-disciplined and human focused. (Tellingly, he was educated in interior design and architecture.)" - I agree, and that is my take as well.
"To suggest that graphic design is solved, is like saying a radio is solved. Technological changes coincide with changes in human understanding, taste, response and feedback." - I think I regret using the word "solved" (hence the quote marks), I think the correct take on this is that there are no major gaps to fill in International style. Vignelli, Paul Rand, Unimark International, Eliot Noyes, and countless others - they all explored the contours of design but only to realize that principles of International style serve almost every purpose and will continue to do so in future - with new technology, the principles don't change. Its application changes. Even today, design houses such as C&G&H continues to excel whist Pentagram is flopping around in New York trends. Paula Sher is an absolute disaster (See LOC redesign by Pentagram).
I don't mind good, well thought out constructive criticism, I would appreciate some more insight into your claims. "I love it when people make authoritative statements about design. Because everyone has an opinion about design, it must mean everyone is expert." is not the tone I'd like to see on HN. I shall also add that my initial comment was rather tongue-in-cheek rant, I should have been more objective and sound. :-)
Fair cop, though my comment was in direct response to yours. To be honest, the bile against design is an opinion I see voiced quite a lot here and elsewhere and it riles me a bit, so I did not take your comment as tongue in cheek.
>The trend at the time Mid-Century design with some influence from 30 year old Bauhaus movement.
Absolutely! You don't think this was a trend? Art Deco was a trend for 30 or 40 years, Art Nouveau maybe a hundred. The Bauhaus was an expression of renewal against the old world of decorative arts, and for a functional, scientific aesthetic. But maybe only true to this until Mies took over.. and he was more an aesthete than functionalist, and perhaps the one responsible for its dissemination across the world as the international style - at least in architecture. Anyway, point was the Rams trend was a revival of the Bauhaus style, clean, modern, healthy, wealthy, and white, against the skeumorphic timberwork of his contemporary's electronic product design. All at the height of the popular dissemination of modernism to the world.
>I think the correct take on this is that there are no major gaps to fill in International style.
I took your use of "solved" to mean precisely this. I just disagree. Perhaps I am a little biased by my field (international style was less successful for architecture than graphic design, though far more widespread).
The thing is, sure, design is partially about the possibilities for composition afforded by the medium and the physical, ergonomic and responsive characteristics of the user, but it is mostly about humanity. A chaotic mess of social feedback effects, a constantly shifting base of understanding and recognition, a wide variety of sizes, shapes, outlooks and interaction patterns. Most of design is manipulating an unseen virtual world of the senses, experiences, social interactions and virtuality in the minds of the users. Small wonder when faced with the scale of the actual problem (not to mention the leanness of economic incentives), lesser designers occasionally lose track of some of the harder rules of design. Even excluding the issue of taste, or as I like to think of it, prejudice ;)
The 'Stripe gradient' on their home page, and a few other things are purely aesthetic, and frankly extraneous to the more functional aspects.
Most of the 'guidelines' from Material were reasonably pragmatic, although very picky, beyond what is necessary in all but the most comprehensive apps with big budgets.
There's nothing wrong with some aesthetic on a home page, and such things will always be trendy and evolving.
Design has two layers.
The first one is based on science. Ergonomics (i.e. edge of the screen is fast to reach with a mouse, color contrast, size of fonts, ...).
The second layer is style. In the given constraint of color contrast, you use _these_ colors, ...
The minimalist style often acts like it is on the first layer, but it is just style on top of ergonomics. It is a good design system. Because it is minimal, it is easier to not make first-layer-mistakes. It is timeless because the second layer is so thin as to not show the trends of the century.
A lot of people think this is good - but I don't think it universally is. As long as you get the first layer right, being more wild on the second layer adds character. Being timeless can also be interpreted as being boring.
Particularly the following comment:
"In some scientific fields/cultures, a stylish website could be viewed as unnecessary or even pompous. In this view, the textual content of a website is the only thing that matters, and if you "need" to make your website stylish perhaps it lacks real substance. This is the same line of thought that supports simplicity in presentation with minimal graphics. I have encountered this especially in math and theoretical CS."
But - and get this - this is an aesthetic in and of itself! It's part of the subtle 'self branding' of academics!
Banking culture, academic culture, SV startup culture, punk culture - they all have their idioms, behaviours, signalling etc. in fact far more so than we'd ever like to admit.
It's more basic than that - the premise of UI deals first core usability, the layout of information of components in a manner that communicates information and allows for basic interactivity, and then to make it all coherent across the app.
That alone can get pretty hard, there's all sorts of decisions that need to be made. Consider that it's as much about what you don't see as those elements you do.
Then you can get into the more granular aspects such as those covered in this article that I identified as 'picky' (for example, the specific contrast ratios among various elements - this is far beyond what 99% of UI designers have the luxury of addressing) and then of course broadening the scope of accessibility, corner functions, cross platform consistency.
Even the basics of UI can be pretty hard, let alone the aesthetic elements, and very specific things.
And of course, all of it 'cross cuts' with the aesthetic aspects, which is why the disciplines of UI and classical design are so tightly interwoven.
Design and UX is very poorly understood, and it's difficult to professionalize or measure given the abstract nature of the subject. As Engineers, we love to measure things, and much of great design defies the kinds of classical measurements we are used to. As the OP indicated, it's also a field wherein it's hard to be self aware i.e. to 'know what you don't know', and so 'everyone has an opinion'.
A great UI is almost like a great API - it's hard to measure and articulate, but generally we know it when we use it.
Industrial design is a different question, but nothing is ever "pure" industrial design. The reason you think Rams's designs won't ever be outdated is down to personal taste and what part of the cycle we're in. Apple dominates the design zeitgeist and thus Rams is fashionable. That too will fade, notwithstanding the validity of his design principles.
When branding hits hammers, I think I might quit the human race. Designers fundamentally do not understand that software is a tool that people use to get things done. Constant reorganization and redesign for trends literally robs people of their skills in learning to use software to accomplish their goals.
I've been using computers since the early 90s. Essentially none of the things I learned early on apply anymore. Fuck, we don't even have menus that explain to us in native language what actions are available to us.
I hate UIs and I hate UI designers. Openly. I need zero innovation in the design of the menus for a word processor or image manipulation tool. I am sick to death of learning this iteration of a file manager or settings menu, dock, or toolbar. And god damn all the icons, they are freaking Greek to me, and if they don't come with a tooltip I am loathe to click on anything that will just go boom. UIs are in a death spiral.
But emacs works pretty much the same. I think I might stick with that another 20 years.
I find this point of view - that designers are inferior to developers - to be incredibly condescending and yet very typical of HN. I'm a developer, and I would hate to work with you or someone who holds this view point. You seem like poison in a development team.
Regardless of whether you meant or not, that's what I see here. A bunch of developers telling designers that their bad at their job, that the devs know better than them, and devs know how to tell designers to do their jobs.
> I am sick to death of learning this iteration of a file manager or settings menu, dock, or toolbar.
You're literally just making stuff up. Apple, probably the most infamous capricious design company, hasn't changed their File manager, settings menu, dock, or toolbar in 19 years.
And no HN people are nice, it's just a matter of understanding the subtilities we might all disagree on.
Ps : a bas php.
What OP is talking about is how sometimes there is bad design, in exactly the same way there is bad code, and using that to extrapolate out to an entire discipline. Imagine seeing someone's shitty very first Visual Basic project and just declaring that all developers fundamentally don't understand their jobs.
It's not about skills, it's about goals.
There are roughly two main conflicting goals when building a product. One, make a product that delivers value to users. Two, make money off that product. In a perfect world, these two would be maximized simultaneously - leading to fair exchange of value between the customer and the provider. In the real world, people sacrifice the first in order to get more of the second.
UI/UX design as it is today is almost entirely about sacrificing utility to boost sales. All the trends parent commenters are complaining about exist to convince users to spend money by means other than delivering a useful product to them - be it by differentiation from others, by looking like popular players, by reducing functionality to make it look simpler when purchasing decision is made. Addiction-inducing design trends like timelines and notifications and loot boxes are a part of this too.
This is not failure of intelligence of individuals, this is a moral failure of the industry. And being on receiving end, as a software user, it's fucking annoying. So is being upstream from it, as a software developer doing their best to deliver value to user, only to see it shat upon because the organization adopts user-hostile design trends.
I think this statement is just about as true for design as it is for development.
That's not to say developers don't make choices sacrificing utility for sales - they do, and I bitch about this here plenty (for instance: bloat, invasive telemetry, privacy-hostile solutions, addiction-generating mechanics). But it doesn't look like a trend that's essentially consumed the discipline.
I don't think this is true. Many undesirable qualities in modern design are due to a cult like belief in them by designers rather than just listening to users. Excessive use of white space in virtually all design being the most obvious one.
That's not generally because it boosts sales. It's because they've really drunk the kool aid that it's better.
IIRC, negative space is one of those things that lay users tend to think is unimportant or even undesirable as a design element, but every empirical study of usability/reasability shows that it is important, and that lacking it produces negative reaction to the end product even if users don't attribute the reaction to the missing negative space, and that this result has held up with online design but was well-known even in older studies of print design.
(This is not an invented example; when I was building a PC recently, I wrote userstyles to remove most of the negative space in UI of a shop I was buying parts from, just to not be driven to madness when trying to compare prices and features among available components.)
Another view on this is that a new set of designers suddenly deciding that the old set of designers were bad at their job and just doing it over can end up pretty terrible for users. Some stuff doesn't need to be disrupted, dammit!
My comment was ranty, but I did not say that.
But fundamentally, it is not about who is inferior. It's about who's subjecting whom to constant change, and who is not listening to whom.
The problem is that it isn't "the designers" versus the developers anyway. We're all a little of both. The problem is that the sum total of all the little changes, both cosmetic and fundamental causes a drift over time. You don't see it until the timescale is long enough. Eventually you drift so far that nothing is familiar. (case in point: last time I used Windows was Win2k. Windows 10 is nothing like it. I have no clue how to do anything, and it is hard as fuck to learn, because there isn't much in the way of all that ugly "text" to help me out. There's no manual, no printed documentation. You gotta google everything, or use the built in searches. )
I find all this really frustrating. I'm worse than a novice. I used to be good at this. Now I'm a dinosaur. I literally cannot use the tools that I was good at 10-20 years ago. While change is constant in modern life, software, and particularly The Web, have accelerated change to a point where we are, as I say in my OP, robbing ourselves of skillsets when using software as a tool.
I mean, talk to someone who used Excel for 15 years, like friends of mine in marketing positions. Every revision. Changes. New functionality, but scrambling the old. And then one day their corporate overloads mandate a switch to Google docs, which not only has way less functionality, but is also all different. It's constantly changing too. The damn thing scrambles itself every 6 months. Eventually, the unconscious brain unlearns what it knew. And, worse it learns that how it is now is not how it will be. The brain won't bother learning anymore. So we stop actually becoming experts. Because what's the point if it's gonna get scrambled in < 2 years anyway?
> You're literally just making stuff up. Apple, probably the most infamous capricious design company, hasn't changed their File manager, settings menu, dock, or toolbar in 19 years.
Mac is relatively stable in the core things (thank god for that!) but Windows definitely not, Linux definitely not, and the web is a total other animal. Mac also drifts a lot. Look at iTunes. But the one MacOS example in a sea of drifting, changing, usually breaking and getting worse, crapass apps isn't really worth much.
When a developer adds a feature to an application, he/she generally reaches for the boring stock widgets in the UI library. It's a full time job just wiring together these components! You end up with an app that looks more or less like every other app. Good for usability, maybe bad for aesthetics.
When a designer adds a feature to an application, their mind is freed from the constraints of the toolbox. Their job is to imagine a better way! Their only constraint is Photoshop. You get... new ideas.
This is of course a cartoonish illustration of these roles, which overlap considerably. But I think there's a kernel of truth to it.
I've worked at several places where the designers made a big show of interviewing/looking at data, but then could not incorporate any aspect of it into their designs. Unsurprisingly, their tools weren't used or appreciated.
The parent didn't make that point.
It's exactly the same with marketers. Or managers. Or anyone who isn't a developer for that matter.
Also 'UI design' and 'design' are separate disciplines, often conflated into the same thing.
'UI designers' 100% understand the nature of 'usability' it's the core of their job - they provide for the actual experience that users will encounter, and it's very hard.
Having done quite a lot of UI design and software, I think the former might be hard in many cases - there are just so many intangibles, and everything cross-cuts making coherency often difficult, moreover 'there are no perfect answers'.
Consider the Google home page, and how many features actually exist in Google search that the vast majority of people are not aware of. The 'google search bar' is the 'most used UI in the world' and my gosh do they have to worry about how to extend that to the range of 'barely literate' people, to regular users, to power-searchers.
Consider the other 'most popular UI': Windows. The Windows 10 release was a fiasco, consider that the richest company in the world struggled to map to a new UI paradigm for a variety of reasons.
At least with software we tend to derive an objective, and we 'solve the problem'. We can say 'it's done' (to a certain degree) and even test for completeness.
And why they're not aware of them? Google does an excellent job of hiding features sought by users (and then sunsetting them when they're not used; I wonder why they're not used). The design there does not serve the users' needs.
Hammers were probably some of the very first things 'branded'.
You want a quality, reliable, robust tool, and surely blacksmiths and makers developed their own reputations for quality among other things.
Branding can be some flakey mumbo jumbo, but it doesn't have to be, in fact in shouldn't be. The three rules of branding are 'authenticity, simplicity and consistency' - and the better brands are good at '1' i.e. 'authenticity', meaning the brand truly reflects the nature of the product or organization.
Yes, this screenshot of MacOS 9 from 1999 clearly bears no resemblance to anything we use today:
Try comparing Windows 98 to Windows 10.
Or some random website from 1999 to that same site now.
I think about design in same way, whether its web design, garden design or industrial design. For example, similar trends have picked up in Cars. RGB mood lighting - everyone is adding it now without fundamentally considering its utilitarian value. Or even if they realize it is "decoration", market forces and consumer choices are obliging car manufacturers to install RGB interior lighting. Same thing with brake light bars that span the entire width of the car. Even 2020 Porsche 911 had to add one!
I know a lot of people think Material=Professional. But I think Material looks more like a small business using a drag-and-drop home page builder. Yes, that goes for the Android interface, too.
I never understood the hatred for skeuomorphism. It's intuitive. It's pretty. It's colorful. It's artistic. A bunch of stacked squares isn't any of those things.
Fortunately, I still have a launch day iPhone that I occasionally use for music, and it can't be "upgraded" out of its skeuomorphic splendor.
That is probably correlated with my taste for minimalist/stylized animation/graphics.
Arguably often it appears just low effort, empty and forgettable.
Opinion: I also find that, while each Material component seems intuitive in isolation, systems layering many if them together usually end up very difficult to use.
I cannot find my way around YouTube creator after all these years.
The problem is that they are consistently bad. Web design has stagnated because of Material. There's little innovation going on because Material is "good enough" and "everyone else is doing it." That breeds mediocrity, not advancement.
For lack of better (less charged) phrasing, it's revolutionary.
I personally coded way more table based drop shadows, tickers, and splash screens than I wanted to. Those ideas then, just as now, came from lazy product managers, small budgets, and general mediocrity.
In 2025, we will see a resurgence of GNU.org style websites. Times New Roman fonts, no css only tables, <3kb markup sites. Baskerville is going to make a huge comeback! Most code will be in Computer-Modern fonts and dot matrix printers will be prized on eBay. Future is gonna be fricking amazing! We are going to have APOD style posters!
On top of that you have so many free templates and drag and drop website designers, as a designer you can’t really afford making good design unless your customer is a multinational company, because customers will always compare you with the cheap online service. This means most designers are basically forced to use something fast, that looks halfway decent by economic reasons. Of course there are always those who really think what they do is great, but I don’t think they are in the majority.
That beeing said, I love the interfaces of old professional electronic gear (and lately I have been working in some myself). I don’t think you can transfer that in a good way to the flat screen without adjustments. Let’s not forget the abnominations bad skeuomorphic design produced, every trend will consist of a lot of cheap trash always.
This hurts on a personal level. While it is a very convenient thing if you are a solo developer, it can easily become a pain on a team, where some engineers care about code quality noticeably less than others and non-developers (e.g., PMs) get involved.
This rarely ever happens if you are a developer interacting with similar people (e.g. as a freelancer). They think you are some kind of magician. Similar when you are a good audio engineer or something, they will just let you do your thing.
The funny thing is: this perception is completely independent of the quality of your work. They will always have something to say – sometimes even stuff that makes sense, but very often just for the sake of it, with no real reason. Everybody thinks they have a good taste and a great eye and what kind of graphics are needed, while in fact often they have none of the above.
To be fair, I think you see the same thing with "software engineers". Many don't have any interest in disciplined engineering or the principles of computer science, they just want to write code.
They also save power on OLED as the guidelines and other commenters have mentioned.
And OLED surprisingly saves a lot of power on Dark Mode, so it's kind of our duty to follow suit.
The saddest part of the (pseudo-)loss of the skeuomorphic design was the fact that many very talented designers found themselves underused, tweaking gradients, while their icon design was an art.
This makes about as much sense as saying music was ‘solved’ by Radiohead or Nirvana. Trends change as culture and tastes change. People deride ‘hipsterism’ but without them we would all be drinking lousy beer, eating crappy burgers and drinking low quality coffee. Trends change, challenge what’s gone before and make things nicer.
My rant is rather a lofty take on the entire culture of design. I had to let it out. :-)
Huh, really? I'm still seeing Bootstrap everywhere. I wish I was seeing Stripe everywhere.
Can you give some examples of Stripe-inspired design? I'm curious to see some examples of execution of that.
"Is it just me or do all new tech websites look like stripe?"
Here is a quick run down: https://www.pages.xyz/
A documentary touching that subject is Helvetica . It is about the Helvetica font.
Seriously though, I like the Swiss/International aesthetic, are there any CSS frameworks / design systems / whatever that implement it?
Love this page in their docs. I constantly refer to it when I need to think about colors even in non-web/technical endeavors
It's like saying you like a pop song that broke through, it's okay to like a pop song. But then you have to live through a dearth of similar pop songs over and over again for awhile.
He also made a typo, he meant Bulma CSS.
I disagree, it's quite obvious they were designed in the 50s/60s.
No way you can say this is 50’s, can you? If so, what gives it away?
 https://www.myfonts.com/fonts/castletype/standard/medium/ according to Whatthefont
these would make for a kickass site and app designs
Kinda fun to think about all the subtle clues we're giving off without even realizing it. Of course any of these observations could lead to a totally wrong conclusion, which makes me start to think of all the fun ways you could "disguise" yourself in modern society.
here is a peg grammar:
text = $.*
edit: sry to disappoint but this matches binary files too :(
Perhaps instead the UI could just slightly bump everything from black to the darkest gray, until the scrolling has finished?
> "On OLED screens, turning pixels on and off can cause a delay when the screen is scrolled, making the pixels blur."
Win some, lose some.
> UIs that require efficient battery usage can use true black. In these cases, some devices (such as wearables with OLED screens) can turn off any pixels that display black to conserve battery power.
Regardless of specific hardware implementations, it can be agreed upon that patterns which can cause some non-negligible amount devices to be less energy efficient are anti-patterns.
I'm pretty sure it works a little differently, but that's the idea.
Why is light UI the default if dark theme confers so many advantages?
I know people that prefer light themes for very reasonable aesthetic reasons (dark themes can seem "moody" or "gothic" to some), but I've still convinced some of them on the battery life arguments.
It's also more effort to develop a dark theme, in my experience. Most frameworks and project templates just start you off with a light theme.
It reminds me of old vector graphic games like Tempest. I don't know much about design, but I feel like it's "harder" to contrast with black than white, and so the bright lines of color against the black background actually give me happy feelings. Weird, but I genuinely think it improves my coding experience.
I like it so much I've been slowly porting the theme to Emacs: https://github.com/komali2/Emacs-VSCode-Default-High-Contras...
When I have some spare time I'd like to sit down and actually learn more about design concepts around these kinds of dark high contrast visualizations, and old vector graphics. Ideally I'd like to learn how to incorporate them into my own web and game design... one day!
Never knew this, but I do recall seeing the blur when scrolling iPhone Xs in-store.
Mixed with root variables
But it seems that “prefers-color-scheme” is still, well, preferred .
ReferenceError: Can't find variable: IntersectionObserver
Which I think is ironic.
Because if you see just that icon it's really hard to recognize what it is, I bet many won't see a ship it (at least I didn't). Only because of a realistic ship icon shown next to it that one realizes that it is in fact a ship.
This is a strange choice, considering that the guide has a very strong discouragement against emulating shadows…
> In a dark theme, components retain the same default elevation levels and shadows as components in lighter themes
It looks pretty good in the examples they give IMO. Considering that the default makes such heavy use of elevation it makes sense to keep that concept in the UI.
> Default themes use shadows to express elevation, while a dark theme also expresses elevation by adjusting the surface
What works for me in this case is a Sass mixin which accepts a layer as an integer in the range 1-N and casts a shadow based on that layer and the total depth N. Consistency is key.
It's almost as if Apple beat Google with their Mojave dark theme with shades of green. Now the material design spec cannot be associated with the same dark / green... except for the plus thumb button on the bottom left.
What this means is there should be an easily accessible, systemwide toggle e.g. Android's quicksettings that asks apps and websites for the theme the user wants at that moment. This is something I've wanted for years, and I'm surprised I haven't seen more people calling for it.