Hook: In today's Hacker News / Lobsters / Habr digest, a link appeared to Marcin Wichary, Unsung: "Deeper dive: Were Touch Bar's problems software rather than hardware?" — an article published in April 2026, exactly ten years after the Touch Bar's release (November 2016 → summer 2026). I scrolled past, read the first paragraph — and stopped. Because Wichary does something nobody else did: he dissects the Touch Bar not as a failure (everyone already did that) and not as a breakthrough (only Jony Ive did that in 2016), but as an engineering artifact that had the right intentions, the wrong level of abstraction, and an absolutely predictable end. And here's what really hooked me: this story turned out to be architecturally identical to two other Apple failures — the butterfly keyboard (2015-2019, 4 generations, 5 years, $50 million class action settlement) and the Magic Mouse with bottom charging (2015-2023, 8 years). Apple three times in a decade stepped on the same rake: took a strong idea, implemented it in hardware before it was ready in software, and refused to remake it when it became clear the concept didn't work. Touch Bar isn't about a strip above the keyboard. Touch Bar is about how a company's cultural DNA can spend years defending a mistaken decision from its own users.
To understand the dispute, you need to keep four dates in mind.
October 27, 2016 — Apple unveils the new MacBook Pro with Touch Bar. This is the first Mac where the top row of function keys is replaced by a narrow OLED touch strip (2170×60 pixels, Multi-Touch), with a Touch ID sensor to its right that doubles as a power button. Jony Ive tells CNET: "Touch Bar is the beginning of a very interesting direction." The team spent two years experimenting with large trackpads rich with haptics and concluded that the best of both worlds was a context-dependent interface on top of regular mechanics.
2017–2018 — Mass disappointment. The main complaints from the professional community: no physical Esc (programmers lose finger position, miss the key), no tactile feedback (can't press without looking), poor application support (most apps even two years later don't use Touch Bar). Apple responds with ... nothing. No software updates, no new APIs, no interaction redesigns.
July 2018 — Apple launches a repair program for the butterfly keyboard. A parallel story that matters to our narrative: the company stubbornly refuses to admit that a keyboard where key travel was reduced from 1.5mm to 0.5mm for the sake of 0.4mm thinner chassis collects dust and stops working. Apple will release four generations of the butterfly (2015, 2016, 2018, 2019) before quietly returning to the "scissor" mechanism in the 2019 Magic Keyboard.
2019 — The only hardware update to Touch Bar: the physical Esc key returns (but not on all models). The touch strip becomes slightly faster. And that's it. Apple does nothing else with Touch Bar until discontinuation.
October 2021 — Touch Bar removed from MacBook Pro 14" and 16". These models get a full row of F-keys, physical Esc, and nothing touch-sensitive above the keyboard.
October 2023 — The last model with Touch Bar is discontinued: the 13-inch MacBook Pro (it lasted longest because Apple wanted to keep a "budget" option). Seven years. Seven. Almost as long as the 12-inch MacBook existed (2015-2019), and three times longer than the iPhone 5C.
April 2026 — Wichary publishes his analysis "Were Touch Bar's problems software rather than hardware?" Ten years later, with the perspective of experience, emotional distance, and access to how exactly Apple buried the feature during those years (rather than how it was marketed in 2016).
Wichary writes a long, empathetic, careful piece. His central intuition: Touch Bar failed not because of hardware, not because of software, but because of the gap between them.
The argument unfolds in three layers.
Layer one — the haptics that weren't there. Apple in 2016 already had excellent tactile feedback technology in the Force Touch trackpad. The same technology could have worked in Touch Bar: light touch — soft "ripple," deep press — confident "click." Apple didn't apply it. Why — never explained publicly. Result: press without looking — you miss. Missed — pressed the wrong button. Esc, which should be an escape in critical situations, became the least reliable element of the laptop for touch typists. And this isn't trivial: for a programmer, Esc isn't a button, it's a reflex. Taking away their reflex is like replacing a car's brake with a button that sometimes works.
Layer two — size you can't live in. 60 pixels high. Apple itself in 2007 in iPhone Human Interface Guidelines established minimum touch target size: 44×44 pixels. That's 7mm physical width. Chevrons in Touch Bar were exactly half that — 3.5mm. Apple, the company that spent a decade teaching the industry UX standards, violated its own standard in its own product. And nobody inside the company stopped this, because by 2016 in Cupertino "design over usability" already worked as an unwritten constitution.
Layer three — software that nobody wrote. Apple put Touch Bar at the center of its marketing presentation. And then ... simply didn't update it. In 2017–2023, Apple released not a single significant software release for Touch Bar: no new APIs, no redesigned control strip, no public "here's what good Touch Bar apps should look like." Quick Actions, the only customization feature, remained "extremely raw" (in Wichary's own words) — he tried to configure it through Automator, and at some point discovered that different settings panels don't talk to each other. This is a symptom: Apple didn't invest in Touch Bar what it invested in the trackpad, Multi-Touch on iPhone, or AirPods.
And here's the most damning thing in Wichary's article: there never existed a version of Touch Bar with haptics. There never existed a version of Touch Bar with deep customization. There never existed a version of Touch Bar that the community could turn into something useful for themselves. Apple released a product, gave it 7 years of life, and never let it grow up.
To get beyond "Apple screwed up" (everyone knows that), you need to look at Touch Bar as an architectural solution. And then a structure emerges that explains not only Touch Bar, but also the butterfly keyboard, Magic Mouse bottom charging, and even Apple's refusal to put a touchscreen in Mac.
The architectural argument goes like this: Touch Bar tried to solve an interface problem (contextuality) at the hardware level (fixed location on the chassis), rather than at the software level (open configurable abstraction).
When Apple designed Touch Bar, it had one specific use case in mind: "photographer in Lightroom adjusts exposure slider." This story works. And it works beautifully — as long as Lightroom is the only application you're thinking about. But a laptop's interface isn't Lightroom. It's dozens of different applications, each with its own shortcuts, each with its own state model, and each with a dozen different roles at different times. Touch Bar was positioned as a universal solution but implemented as a set of demo features for specific keynote presentations.
This is the same mistake as in any hardcoded configuration in IT: when you hardcode into hardware what should be in config, you doom the product to death at the first change in requirements. Apple hardcoded into Touch Bar:
Each of these assumptions was wrong on average, but right in 5% of cases — and in those 5% of cases Touch Bar was truly magnificent. This is why the feature couldn't die quickly: it worked too well in narrow scenarios to abandon immediately, but worked too poorly on average to become indispensable.
In April 2021, six months before Apple finally removed Touch Bar, designer Barry Collins published in UX Collective "This is what Apple's Touch Bar should have been" with a subtitle that says it all: "Apple's Touch Bar may be on its way out, but it could have been different if it behaved like the Elgato Stream Deck."
Elgato Stream Deck is an external panel with 15 (or 6, or 32) buttons, each with its own small LCD screen that can display any image and trigger any action. Want a button to launch OBS? Done. Want a button to paste an email template? Done. Want a button to send a specific SQL query to a specific database? Done. The user programs the panel themselves, rather than waiting for Adobe to add Touch Bar support in their next version.
This is the difference between two approaches:
And Stream Deck, which appeared in 2017, still exists (now Elgato has Stream Deck +, with built-in dials and a touch strip, and this is exactly what Touch Bar should have become). Touch Bar doesn't.
This is a fundamental architectural failure. Apple sold a configured device (closed platform) when the market needed a programmable device (open platform). And for 7 years Apple never found the strength to switch Touch Bar to a mode where the user had a say.
Now — the most interesting part. Touch Bar isn't a unique failure. It's the third in 10 years where Apple takes the right idea, packages it wrong, and can't stop.
Failure #1: butterfly keyboard (2015-2019). Apple wanted to make MacBook thinner. The easiest way — reduce key travel from 1.5mm to 0.5mm. Apple developed the "butterfly" — a mechanism where the key is pressed from top and bottom, allowing reduced height. Technically flawless. Practically — catastrophic. The mechanism collected dust, crumbs, and after 6-12 months keys started sticking or working intermittently. Apple released four generations (2015, 2016, 2018, 2019), each with minor improvements (silicone membrane in 2018, for example), but refused to admit the problem was systemic. In May 2019, four years after the first release, Apple finally returned to "scissors" in Magic Keyboard. Meanwhile a class action lawsuit proceeded, concluding in 2022 with a payout of $50 million — that's $50 for every problematic MacBook sold in the US. Seven years from first release to complete abandonment. Seven.
Failure #2: Touch Bar (2016-2023). Apple wanted to make MacBook context-dependent. Implemented it as a touch strip above the keyboard. Technically interesting, practically inconvenient. Seven years from first release to complete discontinuation. Seven.
Failure #3: Magic Mouse with bottom charging (2015-2023). Apple wanted to make the mouse's bottom surface "clean" (no visible port). Moved the Lightning port to the bottom. Result: to charge the mouse, you need to flip it upside down — and it doesn't work during charging. In 2023 Apple finally acknowledged the problem and moved the port to the front edge (USB-C). Eight years.
Three failures. Seven-eight years each. Same company, same corporate culture, same pattern: brilliant idea, poor packaging, years of defending against feedback, quiet abandonment.
This requires an honest conversation, because explaining it through "Apple is arrogant" is a lazy answer that doesn't explain the structure.
The real reason is cultural DNA where the design team has absolute priority. In 2015-2019 Jony Ive led design, whose CNET interview about Touch Bar is pure design thinking in its highest form: "Doing something different is easy. Doing something better is hard. Touch Bar is the beginning of a very interesting direction." In this interview there's not a single sentence about the user. Not one "how this will help a photographer process a wedding shoot faster." Not one "why programmers will accept this." Design exists for design's sake, and in this world "interesting direction" is the highest compliment.
Second factor — lack of feedback within Apple. When Ive wanted MacBook to be thinner, nobody in the company said "this will break the keyboard," because that would mean challenging design dominance. When Phil Schiller showed Touch Bar on stage, nobody raised a hand and said "how will we update this in software?," because that would mean admitting that design without operational support is just a hardware prototype. Apple is a company where the design vertical has no checks and balances for over 25 years. This works when the design team hits the nerve (iPod, iPhone, iPad). It fails catastrophically when the design team goes inward (butterfly, Touch Bar, bottom charging).
Third factor — financial asymmetry of the decision. Apple spends billions on Touch Bar: research, prototypes, production integration, marketing. Admitting Touch Bar is a mistake two years after release means writing off those billions. It's much cheaper to "support" the feature for four more years, gradually lowering priority, and quietly remove it when a new generation of laptops with a different processor comes out. This is the same logic by which NASA didn't cancel the James Webb Space Telescope after 20 years of delays and 10x budget overrun. If you've already invested reputation and money in an idea, abandoning it costs more than seeing it through, even if "seeing it through" means "pretending everything's fine."
If you take only one lesson from this story, it's this: a good engineering idea without user feedback dies in the same time as a bad one. Time isn't a filter, it just scales the result.
Three specific lessons:
Lesson 1: don't confuse "hard to make" with "needs to be made." Touch Bar was engineering-complex: its own OLED display, separate T1 ARM processor, macOS synchronization, Touch ID, APIs for developers. Apple invested enormous resources, and that's exactly why it couldn't admit the mistake. If Touch Bar had been just a software feature, Apple would have closed it in six months. Complexity is an investment in idea self-preservation, and at Apple this works as a closed loop.
Lesson 2: architectural solutions must be at the right level of abstraction. Touch Bar tried to solve an interface problem (contextuality) at the hardware level. Stream Deck solved the same problem at the software level (programmable buttons). One died, the other lives. Transferring responsibility to the user isn't a product weakness, it's an ecosystem strength. Apple in 2016 wasn't ready to accept this, because it would mean admitting that a Mac user might have an opinion about their own interface.
Lesson three, on the bottom. There's a difference between "Apple failed" and "Touch Bar failed." Apple as a company survived all three of these failures and continued dominating the premium laptop segment. This means UX failure is acceptable risk in a business model where you sell an ecosystem, not a device. Microsoft can afford to fail Zune because it has Office. Apple can afford to fail Touch Bar because it has iPhone. In companies with strong ecosystems, UX experiments become expendable material, and Apple understands this better than anyone. Which doesn't make the story less instructive — but makes it less scary for Apple itself.
If I were designing Touch Bar in 2016, I would do three things differently.
First, I'd give the strip to the user from day one. Not "later, in two years, in macOS 10.14." From day one — SDK, templates, design guidelines, widget store, custom shortcut support through Automator (which already existed). If Apple in 2016 had released a programmable Touch Bar, we'd now be talking not about failure, but about a new industry standard. And Apple would have been ready for Vision Pro much earlier, because it would have had experience creating UI for personal, configurable touch space.
Second, I would never have removed physical Esc. This isn't an ideological question, it's physiology. A programmer presses Esc 200 times a day without looking. If your design requires them to look, you've lost — not to a competitor, but to your own keyboard, on which 30 years of reflex works.
Third, I would have left Touch Bar as an option, not standard. Laptops with and without Touch Bar, at a $50 price difference. Those who like it — buy it. Those who don't — don't. Apple collects data, understands whether the feature is needed, and either develops it or discontinues after 3 years, not 7. The most expensive thing about these failures isn't the failures themselves, but the years during which Apple didn't let the market tell it the truth.
And history teaches us exactly one thing: when was the last time you heard Apple publicly admit a specific feature failed before it was discontinued? If the answer is "never" — that's the root. Design dominance at Apple is so strong that admitting mistakes became a luxury the company can't afford. And as long as this is true, we'll be reading analyses like what Wichary wrote for another ten years — after Apple removes the next feature that "was the beginning of a very interesting direction."