★★★★★ (as books)
★★★★☆ (for the purposes of this blog)

I still remember Mac OS X arriving on the scene with icons that felt infinite in every possible way: in size, in color palette, in dimensionality. We got used to them over the last quarter century, but Michael Flarup’s books rekindled that feeling for me; the icons presented here are lavish, larger than life, and basically pixel-less.

I do not generally like coffee-table books. But I really liked these. The iOS App Icon Book came out in 2022, and the macOS App Icon Book followed two years later. They’re “almost-coffee-table” – which is a compliment! – extremely well-made but portable, and with soul, and thoughtful details, and inspiring evidence of being labours of love.

Each one has an almost-absurd amount of icons (I counted almost 1,200 in one book, and consequently didn’t even attempt counting in the other), but it’s not just the quantity that impresses. The icons are laid out carefully on gorgeous color-coordinated spreads. Many appear in variations so you compare their evolution over the years. Each one is big enough and printed so well you can study it in detail, and I have not noticed one technical flaw in their reproduction.

In addition to beautiful collections of beautiful icons, the book also veers a bit into history, and design advice, and adds ~10 interviews with icon designers each. Those are welcome additions that elevate the books from a boring coffee-table existence, but those are also its weakest parts – although “weakest” in a comparative sense. The things missing for me in the book are: more work in progress and rejected efforts, more specific advice and hard-learned lessons rather than general-interest interviews, a bit more about recognition of icons when reproduced small on screens, and some harder/​cerebral conversations about iconography and its place in the universe.

On the other hand, I know that of all icons it’s app icons that get to be least concerned with semantics and semiotics, as they’re maybe the closest to just pure art and graphic design. I can understand how talking through it all would be an extremely hard task; all of the fantastic icon designers I know personally would struggle with explaining why their output is better than others. It’s possible the extra “left-brain” stuff I want from these books would also make them less desirable for those who just seek visual or artistic inspiration.

Both books are otherwise basically a love letter to app iconography, and awash in memorable details: delightful covers, colour-coordinated ribbon bookmarks, beautiful ex librissen, and a product index and an artist index.

The price – $84 without shipping (they’re printed in Denmark, so for once Europe gets an advantage) – might be a bit of a showstopper. The books are well-made, but you are definitely paying a premium for a short/​bespoke print run. The volumes complement each other well on a shelf, but you’ll do no wrong with getting either one if two is too much for your budget. (There is also a half-price PDF version, if that’s of interest to you, but I cannot vouch for that.)

Just a few hours after writing about floppy disks, I stumbled upon a bona fide floppy icon in the Bluesky’s iOS app, anno domini 2026:

I imagine this, in a nerdy view deep inside settings, might be more of a fun nod, but it made me curious – does Word still use a floppy icon?

Yes, it does! Right next to the icon-less AutoSave toggle, deep within a veritable kowloon walled city of interface elements.

And yet, maybe I should chill with the jokes – NN/Group revisited the save icon in July of last year and surprise! People still understand them.

83% of participants associated the floppy disk icon with saving. […] Another 13% described this object literally with responses such as “disk,” “disc,” or “this is an SD card for storing information.” These responses were not coded as “save,” but still suggest familiarity with the image.

What a fascinating journey! The icon didn’t change at all, but its perception went from being a literal representation of a familiar object, to a skeuomorph once floppies were replaced by hard drives, to then a symbolic representation of physical media in general (a lot of people think it’s an SD card – or perhaps even that floppy disks and SD cards are one and the same), to increasingly just an abstract symbol that represents saving as a concept, registering similarly to the circular arrows for syncing, and an arrow pointing south for downloading.

NN/Group is itself kind of a floppy disk, trying to walk a fine line between their legacy and reinventing themselves. They’re dismissed by many as old-school, academic, boring enterprise software aficionados, relics of a different era. I see some of that and often disagree with them, but I also sometimes appreciate their rigor, reliance on user studies, and outright dismissal of fashion in UI design. I want to revisit their site in more detail and see how I feel about it today, 30 years after Jakob Nielsen’s books rocked my world.

A great post duet from Craig Hockenberry that flew by on Mastodon and clarified something for me:

[For] the extra work to create a custom SF Symbol, our experience is 8-10 hours per symbol. This is also an expert level task: lots of knowledge on how SVG control points work and how to maintain compatibility across different sizes and weights.

If you’re paying a designer to do this, the cost will be somewhere in the $1000-2000 range. For Apple this is an easy cost to absorb, for smaller developers it’s a big “nope”.

And, of course in the Mac menubar (and now iPadOS) you need a lot of them.

Another subtle example of how out of touch Apple Design is with day-to-day development.

So not only is the overiconification of menus in macOS and iPadOS a bad idea, but it’s also expensive. You could make an argument that it would push people into reusing SF Symbols – ergo “consistency” – but that would land better if we haven’t already seen even Apple is struggling with that on their own (previously, previously).

Savage from heliographe.studio on Threads (via Daring Fireball):

I have been wondering the other day why aren’t there more mouse pointer museums and here’s one – Amiga Pointer Archive! (Amiga was a 16-bit home computer especially popular in Europe.)

Doesn’t work so well on mobile, but it’s fun on desktop. I recommend zooming the page to 200%.

Yet another good post by Nikita Prokopov, continuing the theme of icons in Mac OS Tahoe (previously), going into more depth:

In my opinion, Apple took on an impossible task: to add an icon to every menu item. There are just not enough good metaphors to do something like that. ¶ But even if there were, the premise itself is questionable: if everything has an icon, it doesn’t mean users will find what they are looking for faster.

I always liked this kind of an exercise:

There’s a game I like to play to test the quality of the metaphor. Remove the labels and try to guess the meaning. Give it a try:

Also, this must hurt:

Microsoft used to know this.

Nick Heer at the excellent Pixel Envy, commenting on the above post, adds:

This is a gallery of elementary problems. None of this should have shipped if someone with power internally had a critical eye for consistency and detail. If Apple deems it necessary to retain the icons, though I am not sure why it would, it should be treating this post as one giant bug report.

h/t my friends Scott and Ezra.

A funny 12-minute video by Chris Spargo about why traffic signs in the world are standardized only to some extent. This was interesting to me generally in the context of Europe being more iconographic, and America being more “word-y” in their sign design, which extends to devices, keyboards, and (presumably?) software.

The story why [the old STOP sign] got replaced by the American version is also the story why the rest of our signs still look different, and why they probably always will.

A good post by Jim Nielsen about icons in menus (in Tahoe).

This posture lends itself to a practice where designers have an attitude of “I need an icon to fill up this space” instead of an attitude of “Does the addition of a icon here, and the cognitive load of parsing and understanding it, help or hurt how someone would use this menu system?”

It seems a necessary ingredient of introducing icons to menus is thoughtfulness and guidance around when the icons are necessary/​useful and when not.

It doesn’t help that the Tahoe icons seems to mess up indentation. (I haven’t updated to Tahoe and might skip it altogether. Even just the planetary-scale rounded corners are something that feels very broken.)