While I wholeheartedly agree, I suspect the required backgrounds are to create a uniform format between system, where VisionOS requires round icons for more reliable eye tracking.
It seems like every OS got a little harder to use in order to better vibe with VisionOS, the least popular platform they have.
While I applaud the commitment to building a new platform, I don’t like that’s is coming at the expense of the others.
They found having round icons made people look at the center, rather than the edges and corners. Since the UX relies entirely on where the user is looking, this made it more reliable.
I remember reading it in the HIG when VisionOS came out and everyone was complaining about the shape. I went looking to see if there was a reason, and there was.
The thing sticking outside of the icon draws the eye to it, which means your focus is at the edge instead of the center, which makes for a more error prone experience.
Since the eyes are the cursor, this is a problem. Desktop and mobile don’t have this issue.
Regardless, you don't need to make the hitboxes one-to-one with the graphics. Indeed, doing so tends to make for unreliable hitboxes, so most picking systems have two different, idealized enter and exit hitboxes for each icon.
I really disliked previously, when icon prominence could be wildly different because one icon takes up the full area with a big square, while another is a circle that necessarily has a significantly smaller area within the same extent. Icons from Apple were all nicely balanced in size, but third-party apps could be anything.
Giving equal visual weight to each icon is an improvement. iOS was a step forward in this direction, and now they finally brought the same standard to Mac.
Squircles aren't ugly, they're functional. "Shape" hasn't disappeared as a distinct visual cue, as the area within the squircle is made of, well... different shapes.
And let's not forget the fact that Macs still effectively use icon masks. A smaller icon is harder to click, because clicking on a transparent area... doesn't click at all. I remember icons like a skinny letter "S" that you had to click just right or you couldn't at all.
> Giving equal visual weight to each icon is an improvement.
Equal visual weight is another way of saying less differentiated.
> "Shape" hasn't disappeared as a distinct visual cue, as the area within the squircle is made of, well... different shapes.
Shape refers to a boundary outline, not interior patterning. A square with polkadots is still shaped like a square.
> A smaller icon is harder to click, because clicking on a transparent area... doesn't click at all.
That problem is only tangential to what shape they allow your icon to be within an enclosing NxN hitbox. Assume an implied framework where clicking on them isn't broken.
I’m not a designer but I disagree. I want to be able to easily distinguish apps without much focus or concentration or searching. Making them visually distinct with shape and color is superior. Uniformity is a problem not a target.
They've gone too far on enforcing uniformity of icons and abusing liquid glass, but I disagree that arbitrary shapes were better. All the random icon shapes looked cool in isolation, but were harder to scan at a glance. The uniform squircle is a useful constraint.
I wouldn't mind if they allowed something similar to that audio hijack icon, where you require the rounded rect as the guiding frame but are allowed to have some elements protruding out of it. But completely arbitrary shapes are too jarring imo.
I mostly lament the simplification of app icons as an artistic loss, not as a usability loss. Shameless plug, but I made a project based on the idea of icons as pure art with no utility https://www.benedelste.in/post/__001
I can see both sides. Artistic constraints can suck, but on the other hand, for every app with a truly beautiful icon design, like the ones listed in the "It Doesn’t Have to Be Like This" section, how many apps have truly awful icon designs? The dock is prime visual real estate, and as a user, I'd like some kind of constraint that makes it less likely some crazy art style is going to be imposed on my desktop just because I need an app there.
The Liquid Glass slider is an embarrassing outright admission of failure. Apple built its brand as a tastemaker, so to put out this new, controversial design language, and after a year of tweaking, finally throw their hands up and say "we don't know what looks good, you decide" is so disappointing.
That said, all the changes in iOS 27 are such a massive improvement from 26. The first design turnout with Alan Dye gone is making me feel very optimistic of their direction.
Hard agree! Not only is it less fun and less visually appealing to me, I think forcing the uniform squircle everywhere makes it harder (than it used to be) to distinguish one app from another by icon alone.
Alan Dye was hired into Apple in the Jony Ive era when they were launching the first Apple Watch because he came from a fashion/print background. Before Apple really figured out what the Watch was going to be (a health/fitness accessory for your phone) they were going for the "luxury fashion" angle.
Somehow when Ive left, Dye got put in charge of design even though he had zero experience in software design that anyone seems to be aware of. He was criticized for the years following for a lot of bizarre design regressions that were happening across all of Apple's OSes. Then a few months after Dye himself announced Liquid Glass at WWDC last year, he blindsided Apple by accepting a poaching offer from Meta, seemingly because Zuck isn't aware of how untalented the guy is.
Now Stephen Lemay is in charge, who's been at Apple for many years and actually knows stuff about software design. It's said that within the walls of Apple, a lot of people were very happy about the change, and the first showing of design changes we got since then are looking very good for Apple.
Apple CEOs always seem to want to make a splash via hardware (especially since the guy worked in hardware engineering) but it would be nice if an engineer brings more focus on the software as well.
> This time, however, the changes are genuine improvements. Here’s the refined Automator icon, for example
Uh, maybe. Parts of it are certainly slightly sharper in an unimportant way when viewed at normal icon size and not zoomed way in. I'm not sure that it's any better. And if that Automator icon is the exemplar, then any improvement is extremely marginal. My god it's just such a bad icon. Whoever is managing icon design should be extremely ashamed of themselves.
Show anyone the pre-Tahoe Automator icon and ask them what it depicts and why that fits and they'll be able to tell you that it looks like a robot and robots are used in automation and therefore every time they see the little robot they'll think Automator. Ask them what the post-Tahoe icon depicts and why that fits and they'll be able to tell you fuck all because what the fuck even is that supposed to be if you don't already know.
The great thing about the new multi-layer icon format in Golden Gate is that it finally separates an icon's foreground from the background.
So in theory, it opens the door to returning shape-differentiated icons to MacOS if a future display theme (a successor to the poorly-conceived Clear and Tinted themes) allows the background to be minimized while the foreground is emphasized.
What I would love to see, and should now be possible, is a revision of the Clear theme where the squircle is transparent/refractive and the foreground retains its native color.
I was originally excited by the flat design revolution because it appealed to my affinity with uniformity and consistency. But I believe now that I was ignorant and lazy. Bad design still exists within flat style rules, and it has an even worse and cheap feel to it. Meanwhile we've lost whole dimensions of expression.
I honestly disagree with the author when they say that the Golden Gate icons are better than Tahoes. There are more lines, which is literally removing the point of LIQUID glass. It is supposed to be BLURRY, like LIQUID. I get that it is more readable, but are icons meant to be read? There is no text in them, other than Apple TV which is very distinctive. It seems like they just boosted the sharpness of the icons and pushed to production (or I guess it is technically production-beta?) However I do agree with the point of the article. Icons having things outside of the squircle were unique, fun, and interesting.
Someone with 5000 hours design experience needs to make a common icon theme for a few 100 GTK and QT libraries and standard-names. It feels like it's 1000s of hours of work. And then you have to make them available in a few formats, HDPI, maybe a build system, etc. there are a number of themes but the ones I try seem to be missing one or more of the icons from the set. Just need the right volunteers to build them, and also get a bunch of app-builders to adopt them, and figure out what colour the bike-shed should be (blue).
Why volunteer? Why not find a way to pay someone for the value of their time at market rate and release the product of their labor under a permissive license?
For the last 25+ years I've been hearing folks say: "why not find a way". But then not suggest anything more than that obvious answer.
Please suggest an actual path forward, an actual plan that is more than just "figure it out". And the plan needs to address at least 1/2 of the points I made above.
It's a "Hard Problem". The answer needs lots of time, likely money and at least two humans with strong drive to fix the problem.
As a bit of a shameless plug, I did some in the past[0] and am working sporadically on a "fork" of those[1] but it's a whole full-time work. There are hundreds of icons to do for apps alone. Each one needs to be done in 16x16, 22x22, 32x32, 64x64, 128x128 and 256x256 so if say you have 150 icons to do for apps, you actually will need to do 900 icons. And add to that that you'll need to cover categories, places, filetypes, actions...
Granted, you can do a 256x256 and scale it down to 128x128, for example, but if you care for quality some details will be lost anyway. So that's why nowadays you'll see most icon themes are just a bunch of logos plastered over a shaped background.
And what irked me the most was that a few weeks after that I released that first set via deviantart and opendesktop.org there were websites that included them in their sets and made them available for download in their websites, not even a redirect to my deviantart or opendesktop pages or something. And found out after that that some people were using them in commercial projects and stuff so I had to chase them asking to not use them since they were cc-by-nc'ed.
Never got a single cent of any of that. I love making icons, at some point I was even working for the icons that would eventually become the Breeze set for KDE5 with their VDG, but it happens that I also need money to buy the beans.
I think illustration isn't something too much in the mindshare of open source, so overall support for it isn't great. IMO this has contributed to it. The industry standard tools are all closed, with closed formats, so it just sounds like much more of a hassle vs contributing code/text.
I mean this throughout the whole process. The only standard illustration file format I can think of is SVG, but it's largely a format to export to, not one industry-standard software uses as it's main persistence format.
So for starters, contributors tend to need access to speciality software they probably don't have installed to view and edit the source of truth. This also means you're handling at least two files in your VCS, the closed format acting effectively as a blob, no diffs, etc. and an export file (usually more, for different scales) to actually interface with the rest of the ecosystem; this is the file everyone can open, inspect and compare, the one your build consumes, etc.
This already would be a good amount of friction for someone familiar with the tools, but designers are not necessarily familiar with git, the PR process, etc. Add to it that icons are more subjective than code, which overall should follow certain rules and either works or doesn't, and it overall seems not worth it for a casual contributor.
I agree, but I'm surprised there was no mention of contrast or proposal to restrict colors.
Their first good example bumped up the color contrast. The orange examples in their set of "gorgeous app icons" are just as bad as the slack vs photos example.
I would love if the OS had an option to automatically convert every app icon to greyscale and required a minimum color contrast ratio for the original. Then, the user can pick their own overlay colors (similar to the color tags in finder).
We will never know if this was AI generated or not, but I have started to flinch at this sentence structure.
It seems like every OS got a little harder to use in order to better vibe with VisionOS, the least popular platform they have.
While I applaud the commitment to building a new platform, I don’t like that’s is coming at the expense of the others.
I'm confused here. What do you think is the relationship between round icons and eye tracking?
I remember reading it in the HIG when VisionOS came out and everyone was complaining about the shape. I went looking to see if there was a reason, and there was.
Since the eyes are the cursor, this is a problem. Desktop and mobile don’t have this issue.
I really disliked previously, when icon prominence could be wildly different because one icon takes up the full area with a big square, while another is a circle that necessarily has a significantly smaller area within the same extent. Icons from Apple were all nicely balanced in size, but third-party apps could be anything.
Giving equal visual weight to each icon is an improvement. iOS was a step forward in this direction, and now they finally brought the same standard to Mac.
Squircles aren't ugly, they're functional. "Shape" hasn't disappeared as a distinct visual cue, as the area within the squircle is made of, well... different shapes.
And let's not forget the fact that Macs still effectively use icon masks. A smaller icon is harder to click, because clicking on a transparent area... doesn't click at all. I remember icons like a skinny letter "S" that you had to click just right or you couldn't at all.
Equal visual weight is another way of saying less differentiated.
> "Shape" hasn't disappeared as a distinct visual cue, as the area within the squircle is made of, well... different shapes.
Shape refers to a boundary outline, not interior patterning. A square with polkadots is still shaped like a square.
> A smaller icon is harder to click, because clicking on a transparent area... doesn't click at all.
That problem is only tangential to what shape they allow your icon to be within an enclosing NxN hitbox. Assume an implied framework where clicking on them isn't broken.
I wouldn't mind if they allowed something similar to that audio hijack icon, where you require the rounded rect as the guiding frame but are allowed to have some elements protruding out of it. But completely arbitrary shapes are too jarring imo.
The Liquid Glass slider is an embarrassing outright admission of failure. Apple built its brand as a tastemaker, so to put out this new, controversial design language, and after a year of tweaking, finally throw their hands up and say "we don't know what looks good, you decide" is so disappointing.
That said, all the changes in iOS 27 are such a massive improvement from 26. The first design turnout with Alan Dye gone is making me feel very optimistic of their direction.
If they do, I'll consider upgrading both OS and laptop, but right now I'm holding on to Sequoia
Edit: It'll always be Mac OS X to me, not macOS.
Somehow when Ive left, Dye got put in charge of design even though he had zero experience in software design that anyone seems to be aware of. He was criticized for the years following for a lot of bizarre design regressions that were happening across all of Apple's OSes. Then a few months after Dye himself announced Liquid Glass at WWDC last year, he blindsided Apple by accepting a poaching offer from Meta, seemingly because Zuck isn't aware of how untalented the guy is.
Now Stephen Lemay is in charge, who's been at Apple for many years and actually knows stuff about software design. It's said that within the walls of Apple, a lot of people were very happy about the change, and the first showing of design changes we got since then are looking very good for Apple.
Uh, maybe. Parts of it are certainly slightly sharper in an unimportant way when viewed at normal icon size and not zoomed way in. I'm not sure that it's any better. And if that Automator icon is the exemplar, then any improvement is extremely marginal. My god it's just such a bad icon. Whoever is managing icon design should be extremely ashamed of themselves.
Show anyone the pre-Tahoe Automator icon and ask them what it depicts and why that fits and they'll be able to tell you that it looks like a robot and robots are used in automation and therefore every time they see the little robot they'll think Automator. Ask them what the post-Tahoe icon depicts and why that fits and they'll be able to tell you fuck all because what the fuck even is that supposed to be if you don't already know.
So in theory, it opens the door to returning shape-differentiated icons to MacOS if a future display theme (a successor to the poorly-conceived Clear and Tinted themes) allows the background to be minimized while the foreground is emphasized.
What I would love to see, and should now be possible, is a revision of the Clear theme where the squircle is transparent/refractive and the foreground retains its native color.
Apple spent millions convincing everyone they needed Retina displays and then churned out an update that made them all blurry.
Make the icons be Free on Free OSes like Linux.
Please suggest an actual path forward, an actual plan that is more than just "figure it out". And the plan needs to address at least 1/2 of the points I made above.
It's a "Hard Problem". The answer needs lots of time, likely money and at least two humans with strong drive to fix the problem.
Granted, you can do a 256x256 and scale it down to 128x128, for example, but if you care for quality some details will be lost anyway. So that's why nowadays you'll see most icon themes are just a bunch of logos plastered over a shaped background.
And what irked me the most was that a few weeks after that I released that first set via deviantart and opendesktop.org there were websites that included them in their sets and made them available for download in their websites, not even a redirect to my deviantart or opendesktop pages or something. And found out after that that some people were using them in commercial projects and stuff so I had to chase them asking to not use them since they were cc-by-nc'ed.
Never got a single cent of any of that. I love making icons, at some point I was even working for the icons that would eventually become the Breeze set for KDE5 with their VDG, but it happens that I also need money to buy the beans.
[0] https://miler.codeberg.page/?prj=rekt
[1] https://miler.codeberg.page/?prj=betelgeuse
(They still have different shapes, though)
I mean this throughout the whole process. The only standard illustration file format I can think of is SVG, but it's largely a format to export to, not one industry-standard software uses as it's main persistence format.
So for starters, contributors tend to need access to speciality software they probably don't have installed to view and edit the source of truth. This also means you're handling at least two files in your VCS, the closed format acting effectively as a blob, no diffs, etc. and an export file (usually more, for different scales) to actually interface with the rest of the ecosystem; this is the file everyone can open, inspect and compare, the one your build consumes, etc.
This already would be a good amount of friction for someone familiar with the tools, but designers are not necessarily familiar with git, the PR process, etc. Add to it that icons are more subjective than code, which overall should follow certain rules and either works or doesn't, and it overall seems not worth it for a casual contributor.
Their first good example bumped up the color contrast. The orange examples in their set of "gorgeous app icons" are just as bad as the slack vs photos example.
I would love if the OS had an option to automatically convert every app icon to greyscale and required a minimum color contrast ratio for the original. Then, the user can pick their own overlay colors (similar to the color tags in finder).