The author of this article would’ve loved to claim that he was targeted by Big Rubber while reporting; alas, it was nothing more than a chemical process. It all started with an attempt to restore a Macintosh computer from the early 1990s, during which it was discovered that many of those Macs came with hard drives which turn faulty due to a degrading rubber dampener. Then, as I started working on this piece, rubberized sides of my vintage HP iPAQ handheld were discovered to be cracked. Newer devices were prone to degradation, too: this Christmas, I have inadvertently gifted a Nixon watch with a soft-touch coating which turned oily and sticky.

Rubber does decay—even more so if we’re talking about old electronics and other vintage devices kept as collection items or objects of personal nostalgia. Amateur collectors do not usually have access to the same resources and expertise as specialized conservation institutions, and treating pieces of personal tech as things worth preserving doesn’t span as many generations as with many other artifacts.

With decades of institutional research, it would be strange not to ask professionals for some advice applicable to home collections. 30pin has surveyed several museum professionals and people in academia about preservation of rubber and rubberized materials as applicable to old consumer electronics.

“Disgusting Tar-Like Substance”

Mesoamerican people processed rubber thousands of years ago, and further experimentation by Europeans and Americans led to an explosion of applications in the 19th century. Conservators hence had to devise ways to preserve rubber artifacts well before the modern enthusiast community was formed. In a 1988 study published in AICCM Bulletin, Sharon Blank of the Los Angeles County Museum of Art noted that rubber “is already posing problems in museums,” and listed several conservation practices. More research followed, with some compiled into guidelines—like the one by Canada’s heritage department—for use by accredited museum professionals.

Meanwhile, retrocomputing enthusiasts were forming their own takes on conservation which largely focused on the looks and ignored professional standards. Many people in the scene are still keen on de-yellowing old plastics with homemade peroxide-based gel (dubbed “retrobright”), a process which alters the structure of the material and does not prevent it from turning yellow again. The community is filled with stories which would make a museum curator gasp, such as the one a high-profile YouTuber using a Dremel saw to open a rare computer.

What gives rubber and other elastic polymers their key capability—to return to an original form after being stretched or pushed—is their molecular structure. Building blocks which these polymers are made of are connected by carbon-carbon bonds to each other. When these chains are broken, rubber gets softer, noted Blank in his study; when they are linking together, the material gets more rigid and brittle.

Buttons, grips, and other rubberized components of many old electronics tend to turn into a sticky goo under uncontrolled, real-life conditions. “It seems to happen to different degrees depending on the material, and it crosses many industries,” says Gravis, a US-based pseudonymous vintage electronics enthusiast behind the Cathode Ray Dude YouTube channel. He describes having to deal with the rubber decay on old laptops, PDAs, and weatherproof broadcast cameras: “Sometimes it just becomes unpleasant to touch but doesn’t really leave much on your hands; other times, it’s a disgusting tar-like substance that melts like peanut butter as soon as you touch it and leaves hideous black streaks on your hands that are hard to wash off.” Several mid- to late-2000s devices became sticky after a year or two of keeping them away outside a climate-controlled environment, Gravis notes.

Museums, especially ones which focus on consumer electronics, face the problem on a wider scale. Will Wood, content director of the newly opened Mobile Phone Museum, says they had to recycle some cellphones from their collections because “they are so sticky, and they are so disintegrated.” Registered in the UK, the museum lists over 3,500 devices including duplicates—and, per Wood’s estimates, “about 5 to 10 percent of our phones have some sort of cosmetic issues—with plastic, battery, that kind of thing.”

Plastic degradation can take other forms, too. Dr. libi rose striegl, a Media Archaeology Lab manager at the University of Colorado Boulder, tells that, in a semi-desert climate of the inland state, rubber tends to get hard and crumble. “The age range on our collection is so wide that we encounter a lot of different rubber issues, but the thing that springs immediately to mind is our lie detector from the 1930s that had a lot of natural rubber components,” they say. Beyond getting rigid, these parts emit a distinct odor which striegl described like “the lingering smell you get several days after accidentally melting plastic, plus general mustiness/staleness.”

Preventing (Or Living With) The Decay

“Nothing can be done to halt this decay, but with care it can be slowed considerably,” says Adrian Page-Mitchell, collections officer at the Centre for Computing History in Cambridge, UK. Per the museum’s web site, it preserves over 800 vintage computers, as well as mobile phones, consoles, and memorabilia—about 24,000 items total.

In an email interview with 30pin, the curator advises keeping objects in a dry environment, as stable as possible temperature-wise. “Nothing will hurt an object more than being kept in a hot room, as it accelerates the gassing,” Page-Mitchell notes. He explains that all plastics and rubber emit gas, so items should be put in ventilated bags made of inert plastic and preferably left in the open. “If an object has to be kept in a box for space restrictions, remove it regularly and sit in out for an hour or two, separate all materials from each other,” he advises.

Belts, gaskets, washers and many other internal rubber elements—used in floppy drives or other motorized components—are easy to replace with modern parts, curators agree. striegl notes that vintage computers with built-in printers tend to take non-standard belts for their print drives, though. But there’s no way to replace such details within hard drives without permanently damaging the device, Page-Mitchell says. In cases of a faulty drive, he advises to keep the collection machine stored intact and use modern stand-ins—usually based on a Raspberry Pi computers or Arduino micro-controllers—for the display one.

Cords and cables are, in striegl’s experience, covered in “the worst rubber I run into on a regular basis… always either rigid or sticky, always gross.” They note that rubber bands which some people use to wrap plastic around them decay as well. For that purpose, Page-Mitchell advises using museum-grade string. (He pointed at a Norfolk-based Preservation Equipment Ltd as the centre’s supplier.)

Non-functional internal parts which have gone bad are better be removed, thinks striegl—otherwise, they “might crumble and get in the way of proper functioning.”

Rubber feet were discovered by the Centre for Computing History to react to a wide variety of shelves, including those made plastic, woodchip, and glass. “A project is underway to place all our computers that have them on inert Plastazote pads,” notes Page-Mitchell.

Rather than trying to fix devices which show considerable rubber decay, it’s better to focus on “preserving them in the state which they’re in and making sure they’re not getting any worse,” adds Wood—and his peers share the sentiment. Page-Mitchell discourages any attempts to rework or clean the rubber: “No cleaning can take place, as it would only serve to destroy the object further, no detergents or soaps are ever used on the collection.”

Wood notes that, despite the museum’s desire to have every object in pristine condition, many of them only exist in a degrading one—and people do realize some exhibits were used daily. He admits the challenge is yet to be solved.

“We’ve almost kind of embraced that.”

Modern phones have long diminished the popular relevance of dedicated point-and-shoot cameras or camcorders: they shoot good enough in most scenarios, are always ready, and, crucially, require no purchase beyond what you paid for the phone. But looking at the same market as it originated 15 to 20 years ago is like looking at this statement in a carnival mirror. Back then, a prospective mobile camera owner had to buy it separately and carry it as an accessory—all for the ability to take photos which, frankly, weren’t very good.

PDAs, short for personal digital assistants, lent themselves to this kind of eccentric expansion. While they were marketed mostly as an advanced alternative to paper organizers, their ability to run third-party programs and connect to hardware peripherals made them appealing to power users on the go. That meant there was an audience with enough purchasing power to release a Palm-powered massager, a $200 Bluetooth module, or, indeed, an early plug-in digital camera.

The age of add-on cameras, whether for PDAs or for other mobile electronics, was short even by the early-2000s standards of fast-paced advancements. Today, 30pin looks back at two such peripherals, released few years apart for competing platforms, to show what preceded the feasibility—both technological and economical—to build camera into devices themselves.

Plenty of PDA accessories had a definite and immediate need to fulfill: you bought a foldable keyboard to type on the go, a wireless modem to check you mail at the airport, and so on. But a fair share of them could be filed under “it’s cool we’re able to fit that into your pocket,” with advertised ways to use not keeping up with what technological abilities yet. Early PDA cameras Eyemodule were not inherently more useful than Nintendo’s less “serious” Game Boy Camera. Sure, the PDA accessory could shot photos which you could send and manage as computer-readable files—just like on the real digital camera!—but it was still closer to a toy than a tool.

Sony, a famed gadget maker, showcased a plug-in camera shortly after they entered the PDA space in 2000. It was made specifically for their Clié series of Palm-powered handhelds which earned themselves a lasting reputation over the next six years, with award-winning industrial design and standout features like high-resolution displays and media playback. Sony’s handhelds pushed the limits of what the Palm OS platform was capable of, so it’s a natural target for some hi-tech accessorizing.

The Memory Stick Camera Module, model PEGA-MSC1, was formally announced in October 2001 and rolled into a “Clié Gear” brand of hardware expansions afterwards. The device plugged into Clié’s Memory Stick slot—the same one which, in its many varieties, was the only way Sony would prefer you to expand your mobile storage for years. Michael Khan, marketing manager for Sony’s digital imaging products unit, told the EMedia Magazine in 2002 that, with Memory Stick, Sony “quickly saw beyond storing data and started looking at interlinking devices.” No devices other than Clié handhelds, however, supported these Memory Stick devices or had their own ones.

The module’s rotating sensor with a fixed-focused lens shot was capable of shooting pictures with a maximum resolution of 320 by 240 pixels—the “perfect size for e-mailing or Web posting,” if we were to believe Sony’s press release. Pictures were stored in a proprietary format within Palm OS-specific database file system but could be exported as JPEG files (confusingly named “DCF” by Sony). Despite the Memory Stick appearance, the Clié accessory had no storage of its own, and the camera application had to be installed using a PC.

Pictures taken with a Clié camera had the same horizontal resolution as a screen of a PDA it was designed to be taken with. However, the app reserved only a stamp-sized rectangle in the middle of a screen to a slowly refreshing preview area. Exporting the image wasn’t snappy, either: it took around 30 seconds to convert a single photo to a standard image file, in 30pin’s testing with Sony’s 2002 entry-level PEG-SJ20 handheld and the original Memory Stick card.

Many reviews of the Memory Stick Camera Module focused on the novelty of taking a photo with a handheld computer rather than on the image quality—and, confusingly, that was true for other similar devices throughout the early 2000s. A brief review in a June 2002 issue of Handheld Computing Magazine stated that the accessory “won’t take the place of any half-decent digital camera.” Rob Pegoraro, the Washington Post tech columnist, was far more critical in his February 2002 review: he said, in a rare passing mention, that “the Memory Stick Camera makes for a better demo than a working product.”

It’s not just the size of images, equivalent to a quarter of a resolution of the most basic 1990s monitor, which hurt their quality. Colors on photos takes with the PEGA-MSC1 look like a crude approximation of what the eye was seeing during the capture. Nothing ever feels in focus, and shooting any full-frame object upfront gives the picture a noticeable fish-eye effect. In a controlled environment, a 1997 HP PhotoSmart digital point-and-shoot camera, functionally identical to an older Konica Q-EZ model, produced a much clearer image—the one where you can actually read the logo on a game box, as seen above. The photo taken with an older standalone camera is also devoid of big multicolored blobs of digital noise, found on any Memory Stick Camera Module photo no matter the lighting conditions.

In Less Than Two Years, a Notable Upgrade in Quality

HP released its first plug-in camera for their Jornada line of Windows-based PDAs in 2001, shortly after Sony’s foray. The camera plugged into a Compact Flash slot, had its own viewfinder and a shutter button, and hit the store shelves before HP’s merger with Compaq and the subsequent takeover of the iPAQ handheld series. This article will cover the module’s effective successor, the HP PhotoSmart Mobile Camera, which used an SD Card standard extension instead.

While the PhotoSmart Mobile Camera matched the silver-painted design of iPAQ handhelds it was marketed with, there wasn’t anything which makes the accessory specific to that particular brand of PDAs. Veo, a California-based company, sold the identical plug-in camera without the HP branding and worked with Palm on making a version of the accessory which would’ve carried the Palm badge instead. (The deal fell through, and Palm carried the Veo-branded version on their store instead.)

HP and Palm announced their variants of the same camera within a month of each other, on August and September 2003, respectively. The PhotoSmart was sold at $129.99, according to online discussions. Veo’s web site was listing the same price, but Palm sold it at their online store for the same $99.95 mentioned in the announcement of the unreleased Palm-branded edition.

It’s not clear whether the sub-$100 price point was something Palm and Veo explicitly agreed on and had to honor later. The manufacturer’s site had drivers for both Pocket PC and Palm OS devices, but the latter ones came with no support and “as a courtesy for Palm users.” Veo did not even mention the Palm OS compatibility on the camera’s product page, further adding to the suggestion it was made for the Microsoft platform first.

HP PhotoSmart Mobile Camera shared the general way of work with Sony’s accessory: it, too, plugged into a memory card slot and let you take both selfies and scenes facing the user with a rotating lens. Unlike the Memory Stick accessory, it’s considerably bulkier and, in 30pin’s testing, doesn’t lock into the card slot properly. In spite of shortcomings of the industrial design, the camera takes much crisper photos than a Sony one. While it tends to underexpose most of shots and adds an excessive vignetting effect, it doesn’t turn everything it takes into a rainbow mush like the Sony one.

The indoor picture quality of the PhotoSmart Mobile Camera is below the one offered even by cheapest of contemporary digicams, though, and focusing the shot with a bezel around the lens is tricky considering the minuscule preview area. However, under good lighting conditions, the accessory manages to mostly best its 1996 standalone relative, a goal which the Memory Stick camera was far from reaching.

A look beyond enthusiast web sites and consumer electronic press reveals a peculiar strong point of the PhotoSmart camera: the way it could be integrated with third-party applications, thanks to a development kit made public by Veo. Syware, a Massachusetts-based maker of mobile database software, added support for the camera to their flagship Visual CE product in 2004. The camera found an extensive usage in the mid-2000s research on accessibility tech, object and character recognition, foreshadowing applications like Google Goggles and Google Lens. While later iPAQ handheld had built-in cameras, some researchers noted there was no API for them—in layman’s terms, there was no documented way for third-party applications to get data for the camera—and they had to use the plug-in one.

Adding a Camera Versus Adding Another One

PDAs were not the only portable electronics to support camera accessories. Ericsson and Siemens released similar modules for their mobile phones, and Sony’s video gaming unit was introducing various revisions of a PlayStation Portable camera until 2009. But Palms, iPAQs, and similar devices, due to their positioning as literal pocket computers, both asked for cutting-edge expansions and attracted forward-looking users who appreciated them.

A burst of mid-2010s experimentation in mobile tech led to two concepts which companies like LG and Motorola tried to capitalize on: modular phones and 360-degree photography. Both went nowhere by the end of the decade, with a performance of devices like the LG 360 and Hasselblad’s snap-on zoom camera showing that people were largely disinterested in expanding what they had in their pocket. Consumers were showing progressively more interest in phones which were doing less, forming a market for Nokia’s classic phone revivals and designer devices like the Light Phone.

With the opposition to nuance seemingly embedded into any collective online discussion, there is a risk for any odd class of devices to be called failures. But in retrospective, there is a difference between a step in the wrong direction and a step which organically led to the current state of affairs.

The “Little Tramp” advertising campaign still stands as IBM’s most remembered one—and for all the right reasons. A collection of 1980s TV and print ads, designed to make IBM’s seminal lineup of personal computers more approachable, turned Charlie Chaplin’s most famous visage into the Big Blue’s de facto PC mascot. The Tramp was such an icon for IBM, its ad agency’s chairman had to explain why he wasn’t brought back for the successive campaign. The ad campaign is regularly remembered by professionals, analyzed by scholars, and would surely be recalled in just a few months due to the IBM PC’s 40-year anniversary.

The same can’t be said for IBM’s decision to take the OS/2, their advanced and technically demanding software product, and pay at least $1.5 million a year to share the naming rights with a second-rate, US-centric sports event. The IBM OS/2 Fiesta Bowl, as the event was formally called, united computing columnists and sports commentators in their bewilderment, especially as IBM was bleeding money and losing the a with Microsoft over desktop dominance. It does, however, show IBM’s bold attempt, however misguided, to market the OS/2 among people outside the company’s clientèle.

1990 was a rough year for the Fiesta Bowl. The college football competition, held in Phoenix metro area on a New Year’s Day, was put in a reputational disarray after Arizona voters initially declined adopt Martin Luther King Jr. Day as a paid state holiday. The same year, the bowl lost its first title sponsor, Sunkist Growers, as the citrus cooperative had to recover from a crop-ruining freeze. IBM’s fortunes weren’t great either: the company’s success with the original PC and its successors, the XT and the AT, did not translate into retaining control of the market. One of initiatives which the IBM was trying to build itself back with was the OS/2, a prospective replacement for Microsoft’s technologically limited MS-DOS operating system. IBM and Microsoft originally worked on the OS/2 together, but as the latter was finding its own success with graphical interface of Windows, IBM shouldered the responsibility for the former joint project.

IBM’s arrival as the corporate underwriter for the Fiesta Bowl was announced in December 1992. The deal let the embattled Fiesta Bowl stay in the top league of college football bowls, as IBM’s funds allowed it to hit the $6 million payout goal required by the Bowl Coalition. The three-year Fiesta Bowl sponsorship cost IBM $4.5 million total, according to the International Events Group data quoted by the USA Today. (Another reported figure is $8 million, as stated by a Creamer Dickson Basford employee in Marketing Computers.)

Whatever was the benefit for IBM, it wasn’t as straightforward as a pile of cash—and the company would have benefited from that. The deal was announced amid a series of layoffs, unprecedented for the company which prided itself on a policy of lifetime employment. Billions of dollars were written off to cover befefits for many IBMers as it forced them to retire. The company reported a 1991 annual loss of $2.86 billion, followed by a record $4.97 billion loss in 1992.

IBM representatives defended the deal vigorously. “It would be inappropriate to link one set of business activities with another,” said James C. Reilly, IBM’s US general manager of communications, as he was confronted by the New York Times about the sensibility of cutting jobs and doing high-profile sponsorships at the same time. In another statement, which makes less sense each time you re-read it, IBM press officer Rob Crawley told Open Systems Today that “you need mindshare of the market in order to do better in your market.”

Aside from an assortment of merchandise which pops on eBay to this day, the deal netted the right for IBM to advertise the OS/2 during the bowl as it was broadcast by NBC. A set of “Stay 2’ned for more” ads, which allegedly aired during the first IBM-sponsored Fiesta Bowl, tried to woo the audience with a screen recording of a 32-bit Lotus SmartSuite—whatever that meant for the uninitiated—and such zingers as “why not drag-and-drop yourself out to get it?” The 1994 set, per a short note in the IBM-backed OS/2 Professional, was devoid of computer jargon and more focused on people and activities.

Off the record, IBMers were not impressed with the endeavor. “It should have been called the IBM Bowl,” said an unnamed former IBM employee to Business Marketing in 1994, implying that the name-in-title schema, as opposed to the name-as-title one, did not make people link the competition with its sponsor. Rick Chapman, a former consultant for IBM and an author of In Search of Stupidity: Over 20 Years of High-Tech Marketing, wrote in his book that the sponsorship was known internally as “The Fiasco Bowl.”

Some sports commentators, noting how ridiculous bowl names were becoming, picked the Fiesta Bowl as their object of ridicule. “I am in awe of the IBM OS/2 Fiesta Bowl because I have absolutely no idea what an OS/2 is. For that matter, I have no idea what the OS/1 was,” wrote famed journalist Tony Kornheiser in his December 1993 column for the Washington Post. (Bowl names have evidently become even crazier, giving a fertile soil to many listicles.)

Save for the Arizona Wildcats’s historical shutout loss in the 1994 game, the OS/2 era was relatively uneventful for the Fiesta Bowl. At the same time, IBM underwent a corporate shakeup which set a new course for the company. John Akers, a lifelong IBMer who led the company as it was losing control over the personal computer market, was ousted in a management coup, with Lou Gerstner becoming the new CEO in April 1993. Under Gerstner, IBM hit the brakes on the plan, set in motion by Akers, to split the company into several independent units; cut ties with 40 ad accounts in favor of a consolidated deal; and, crucially, shut down the OS/2 project.

In September 1993, Bob Hunt, the Fiesta Bowl president, accused the IBM party of abandoning a new six-year, $20-million agreement ostensibly backed with letters of intent, says a note in Arizona Republic. The last Fiesta Bowl under the OS/2 banner was held on January 2, 1995, after IBM introduced a new major version of the system to compete with Microsoft’s upcoming Windows 95.

Hunt told that IBM decided to “get out of the sports marketing business,” which didn’t turn out to be that unequivocal. In 1996, for instance, IBM was a title sponsor for an Olympic gymnastics test event in Atlanta, and it continues to sponsor some major competitions to this day—like the upcoming US Open tennis tournament. It did, however, pull out of of sponsoring the Olympic Games in 1998, ending their 40-year partnership with a Sydney 2000 presence. IBM and the International Olympic Committee parted ways with public disagreements over cost sharing and Olympics.com web rights, the Wall Street Journal reported.

The exceptional reception of the Windows 95, a system which codified such common PC user interface concepts as the Start menu, the taskbar, and an array of window controls, pushed the OS/2 out of consumers’ sight. It did get another major version in 1996 and some feature updates till 2001, but as a desktop environment, the system became irrelevant. IBM ended OS/2’s first-party support in 2006, when the system could only be found as a part of legacy industrial projects.

The end of the OS/2 spelled no demise for IBM—or the Fiesta Bowl, for that matter. Gerstner’s management decisions earned him a reputation as a turnaround artist, with books and studies devoted to describing IBM’s newfound direction. His successor, Samuel Palmisano, completed IBM’s exit from consumer computer market by selling its PC division to Lenovo in 2004. These days, the closest IBM has to a consumer operation is The Weather Company’s web site and apps, which are just a public facet of their B2B forecasting business.

The Fiesta Bowl didn’t have to wait long for good times either. While IBM’s exit made the bowl’s managers to scramble for a just-in-case agreement with the household cleaning company Dial Corp, the final deal was signed with Frito-Lay. The snack behemoth, in a match which arguably made more logical sense than the one with IBM, turned the competition to Tostitos Fiesta Bowl. Promoting tortilla chips turned out to be fruitful for the bowl and its participants, as 1996 payouts were increased from $6 million to $26 million.

Tostitos supported the bowl for 18 years—two more years than the OS/2’s entire development process.

Ernie Smith contributed to research.

The way we are choosing what to listen to has changed notably since the age of standalone music players. Modern phones have not only enabled the wider switch to online streaming, but popularized several, different ways to control your playlist, as opposed to a set control scheme of a dedicated piece of hardware. You can now drill down your library with a touch screen, use handset controls to switch tracks, and, if you fancy, rely on a voice as a proverbial control surface.

By 2009, Apple, which still maintained its iPod line of players separately from the touch screen-enabled iPhone, was seemingly confident enough in headphone controls and voice feedback to build a device exclusively around them. The result was a third-generation iPod Shuffle, a diminutive aluminum gadget which had a power switch, a single LED, and nothing else to palpably interact with.

Apple’s smallest variant of its most affordable player was criticized for its elaborate control scheme and, crucially, a reliance on a specific pair of packed-in headphones. The buttonless Shuffle was succeeded with a retread of an earlier hardware version 18 months after its introduction.

In the late 2000s, the iPod Shuffle line was Apple’s representation in the booming market of inexpensive flash-based music players. Costing as little as $49, these players were Apple’s cheapest self-contained products, serving as either a gateway to their wider lineup of devices and services or a cheap enough Mac complementation to buy on a whim.

While competing music players were getting FM radios, voice recorders and other added features, Shuffles were conceptually limited in their functionality. Apple argued that their entry-level players didn’t even need a screen and, to keep your music fresh, should not play tracks in order—hence their name.

Apple marketed the Shuffle as a stylish wearable device from the get-go: Apple’s CEO Steve Jobs, unveiling the first model in 2005, wore it on a lanyard during his Macworld keynote speech. The second-generation model, introduced a year later, was even more of a purported fashion accessory, with a belt clip and a much smaller aluminum case available in ten colors total.

First two Shuffles, despite the work of Apple’s designers, looked undeniably like media devices, with a circular set of playback control buttons resembling the iconic click wheel of more expensive iPods. The only thing which would mark the next Shuffle as one, though, was a headphone jack. The third-generation iPod Shuffle was as small as a USB flash drive, with no features on the front of its metal case.

Playback controls were moved to a three-button remote built into a cord of Apple’s bundled earbuds. To flip through tracks in a linear fashion, you’d use a control scheme which eventually became familiar to any headphone user: click once to play or pause, double-click to go forward, triple-click to go back. But any other action required using a code of short and long presses while relying on audio cues—something you’d miss without reading the manual. The same all-purpose headset button was also used to invoke VoiceOver, a voice guidance which allowed to check what’s playing or, for the first time in the Shuffle line, switch playlists.

In their review of the third-gen Shuffle, Wired called controls “awkward” and “counter-intuitive” whereas iLounge said the remote-based control scheme “completely screws up a control scheme that worked well across two predecessors.” But it wasn’t just the control scheme which annoyed the reviewers, but the need for a specific pair of Apple headphones. The headphone jack did allow plugging any standard headphones in, as the iPod started the playback upon turning on. But with no way to control what’s playing, a stylish music player was as interactive as a wartime fixed-frequency radio.

Control woes led third parties to creating accessories like Scoche’s tapSTICK, a wrap-around iPod Shuffle remote which hid the compact aluminum case behind a shell with a set of buttons. Another solution was Ozaki iCommand Controller, a dongle which extended the player by half of its size while blocking the power switch. None of these plastic remedies, however, offered a full range of controls found on a second-generation Shuffle: just like with Apple’s earbuds, you’d have to rely on a combination of taps to rewind and fast-forward music.

As the third-generation iPod Shuffle lost its mainstream appeal, used goods consumers are finding another caveat: most wired headphones from Apple themselves, including EarPods, are not supported by the player. While Apple’s most recent fitting pair has a similar-looking remote, they do not talk to the no-button Shuffle the way it expects from packed-in phones. Even if you do find the complete third-gen Shuffle kit, there are chances it will come with faulty buds which, per Apple themselves, exhibited unresponsive controls and “unexpected volume increase or decrease.”

With the third-generation iPod Shuffle, Apple built a minimalist device with too many limitations for the core Apple audience. While it would still be years till many similarly controversial design decisions of the 2010s, such as removing function keys from a notebook or limiting the expandability of its power-user desktop, Apple’s penchant for making things smaller and more simplistic was already a source of parody material. “The new iPod Shuffle concept might just fool you for an Onion story: no buttons, proprietary headphones, and next you’re going to tell me it talks to me too, right?” read the Slashgear review.

In 2010, Apple brought media buttons back for the fourth-generation iPod Shuffle. Features of its predecessor were carried over, relegated as an option, not as the only way to control the player. The company went as far in decoupling third-gen features from the packed-in headset as to put a separate VoiceOver button on the player itself. The new model, despite lacking the four-gigabyte variant of its predecessor, continued being sold for the better part of 2010s.

Incidentally, the most recent case of Apple backtracking on its design decisions parallels the Shuffle story. In May 2021, Apple replaced its five-year-old Apple TV remote—a symmetrical black bar with a glass touchpad—with a model which looked like the one which the touch remote replaced. With the new remote, the company made touch controls a part of a tired-and-true navigational button, rather than its own thing demanding a full attention.

On-headset controls, just like touchpad TV ones, were deemed not able to stand on its own. But that doesn’t say they are bad—just not refined or accessible enough to be used as the only input method. These moments in the Apple history show that development does not boil down to coming up with “good” ideas and throwing “poor” ones away: sometimes, a concept just needs to be put in a fitting place.

Correction, June 24, 2021: An early-access version of the story incorrectly stated that third- and fourth-generation iPod Shuffles were released in 2008 and 2009, respectively. We regret the error.

Compact discs were not originally meant to store computer programs. But when they were retrofit to do so in the late 1980s, their storage capabilities triggered a boom which reached software companies, traditional publishers and broadcasters alike. IBM, Microsoft and other were adopting printed materials to the new format and making their own made-for-CD publications. Throughout the 1990s, major media brands, like CNN, Time and National Geographic, appeared on titular CD-ROMs, with many more parties to follow.

The fact you didn’t have to insert a CD to read these words indicates what the fate of the medium was. CD-ROM publications were not destroyed by the Internet overnight, but it left them stripped off innovation by mid-2000s and turned them into a relic by the early 2010s. By now, optical disc drives have long stopped being a de facto integral part of any pre-made computer, and compact discs themselves, partially sold on a promise of durability and long storage life, have shown that the promise was not kept.

Encyclopedias, digital magazines, and other CD datasets became time capsules which represent the knowledge and viewpoints of yore­, say professionals interested in preserving old media. But they also see them as potential sources of research and creative inspiration­­­—and works which need to be not just saved but made once again accessible.

One of appeals of a CD-ROM format upon its introduction was the ability to store arbitrary data—and a lot of it, compared to floppy disks—as long as the information was in a digital, computer-read format. Importantly, CD-ROMs were gaining the foothold during the time when more computers could reasonably display rich graphics, play sounds and video clips. “Multimedia” was the buzzword of the coming decade, and in a few years, compact discs which enabled it “suddenly jumped from the obscure back pages of computer magazines to the front pages of the Wall Street Journal,” as described in 1995 by the late historian Roy Rosenzweig.

It wasn’t just the combination of different formats which made CD-ROM titles stand out, but the way the information was laid out and presented within the software—or, specifically, the fact there was no set way to do so. Most early reference CDs, like the 1987 Microsoft Bookshelf, were modeled after databases and help systems, but authoring software like Apple’s HyperCard and Macromedia Director allowed for a more open-ended approach.

The 1997 Immemory CD-ROM was one of cases where a creator—in this case, a French auteur Chris Marker—used the laser-read canvas to create a non-linear work which, per the retrospective in the Criterion Collection’s Current, let people “perpetually scratch at an ever-expanding surface” of his legacy.

Immemory was made in HyperStudio, a HyperCard-inspired program by Roger Wagner with a greater focus on painting tools. “Chris maxed out this program, using it to develop his full-length CD-ROM, using every bit available to him,” says Damon Krukowski, a musician and co-publisher of the Exact Change imprint which brought Immemory to the English-speaking audience.

The original English version of Immemory was made for Macintosh computers only. As Apple switched Mac’s operating system to a new codebase, it gradually removed components which made titles from 1980s and 1990s compatible with newer computers. In a rare case for the genre, Exact Change remastered the CD-ROM in 2008 to make it work on concurrent Macs, but further updates broke it once again. “It will not open on newer computers. And so far, all emulation programs we have tried have failed,” laments Krukowski in an email to 30pin.

(Centre Pompidou, which publishes Marker’s works in French, released their own, browser-based conversion of Immemory in 2011. Incidentally, the new release was built in Flash, a proprietary web plugin disabled by major browsers in 2016 and discontinued by Adobe in 2021. The Flash version was taken offline.)

“In some cases, the information is only available on that media,” says Seth Anderson, a software preservation program manager at Yale University Library, about CD-ROMs. He notes that the library’s optical disc collection, in addition to multimedia titles, includes conference proceedings, databases and other information which “may still have research value despite its age or format.”

According to the Yale Library website, more than 5,000 discs were once deemed valuable for acquisition—and Anderson’s unit works on making them available once again. To do so, the unit recreates the environment which CDs were supposed to run under, complete with old office programs and media players, and makes it accessible through a web browser. Anderson notes that the variety of CD-ROM materials poses the greatest challenge for the Digital Preservation Unit. “Identifying the dependencies of these discs is not something we can easily automate,” he says during an email interview with 30pin.

The library and its Emulation-as-a-Service Infrastructure (EaaSI) partners make old software accessible by running it on a remote computer, with the web browser serving as a terminal. Another way to put the software on the web is to put an emulator into the browser, with a remote server being used for downloads only. The Internet Archive, a digital library best known for its Wayback Machine project, used this alternative technology, dubbed Emularity, to make Apple II games, Flash animations, and other titles freely accessible from virtually anywhere.

Using the same technology to emulate CD-ROM publications is hard due to the manifold increase in size, says Jason Scott, an Internet Archive archivist. “We have been able to have a million simultaneous users, but some of those users have a terrible experience,” he says. During a Zoom call, he explained than the Archive’s solution downloads the entire image to the visitor’s machine, which then has to be powerful enough to emulate an old system. “If your car is terrible, the show is terrible,” Scott says, as he compared EaaSI and Emularity to a movie theater and a drive-in respectively.

CD formats themselves are not making things easier, says Scott: “Formats which worked in 1989 do not work the same as in 1995. Software that pulls the CD usually knows this, and can usually work with it, but you can be surprised later. There are also bizarre hybrid formats where music tracks and data live on the same disc… Sometimes it would work, sometimes it wouldn’t, and sometimes they just didn’t do what they were supposed to.”

Despite the challenge of emulating CD-ROM titles, Scott finds it important to copy them and make them viewable on the Archive. He says he have personally imaged 5,000 or 6,000 CDs. “Fundamentally, for at least 20 to 25 years, they were perceived as one of methods of providing information services,” he says, adding that “it’s vital, to the best of our abilities, to make as much of it accessible both as its own separate package and as a searchable product.” The Archive’s contributors, including this article’s author, uploaded more than 46,000 software CD-ROMs to its servers.

Compact disc software was always at odds with the connected world. While the early web couldn’t compete with multimedia features of CD-ROM titles, it surpassed them with greater interconnectivity, the sheer wealth of information, and a relatively low cost of access. Dr. libi rose striegl, a Media Archaeology Lab manager at the University of Colorado Boulder, describes that a potential of the CD format was discovered “[just as] we realized we could do basically the same things online with freer access.”

CD-ROM titles couldn’t compete with the web for the ability to present the most recent information to date. Encyclopedias like Microsoft’s Encarta and the CD version of Britannica were published in annual editions, and while they could be updated over the Internet, said updates were nominal compared to what you were getting for buying a new volume. As people were getting rid of last-year editions as they were buying more recent ones, CD-ROM encyclopedias lacked the “inherent need to be kept,” as Scott puts it. He notes, however, that publications—both digital and printed—serve an important function for people researching past time periods “who want to know what was the information in that time.”

Electronic encyclopedias are “like snapshots taken at the time of publishing,” says Tianyu Ge, a graduate student at Virginia Tech. His Master of Fine Arts thesis, Encyclopedia Mundi, draws a direct inspiration from 1990s CD-ROM encyclopedias, complete with user interface elements inspired by old versions of Windows. After showing his work-in-progress on Twitter, he explained to 30pin that Encyclopedia Mundi will be presented to an audience as an unknown piece of software found on an old used PC. “If the audience decides to explore further, they would be able to piece together, among other things, information about the fictional developers,” Ge says in an email. The content of Encyclopedia Mundi, he notes, would be generated by the GPT-2 neural model, to “act against the software [and] subvert the previously established supposition.”

Optical discs themselves—old, fragile, and growingly inaccessible types of media—might have fallen out of widespread use. But striegl, who’s an artist as well as educator, sees a potential interest in the static nature of information which they represented: “The ability to update and correct information afforded by browser-based access is a boon, but having a material thing that could be distributed has been lost.”

Editor’s note: For a 30pin feature on preserving CD-ROM publications, we’ve talked with Jason Scott, an archivist at the Internet Archive. During a lengthy call, he shared his views on preservation issues at large. Considering Scott’s expertise, we’ve decided to additionally publish his remarks on an op-ed basis.

The thing we are more concerned about are mechanisms disappearing. The number one thing that’s very hard to get back, going back 100 years, are machines that read media. People always say, “well, there are millions of them, we’ll be fine.” Another ten years, and another ten years, and suddenly, they are disappearing.

I don’t distinguish anything anymore. I have stacks of floppies and writings and CDs, and I’m imaging video tapes right now. If I can do this right now when the technical process exists, that’s great, because the technical process won’t exist in 20 to 30 years. It would be very, very hard, and it'd have to justify itself.

If you were to rescue something from 40 years ago, you’ll have to say, “Well, it’s gonna cost us $200. Do we really need to do this?” As opposed to right now, when it’s not free, but it’s much less than $200 apiece, and you can do it. That’s where I currently stand on it.

• • •

One of the reasons we are working so hard to make everything available as quickly as possible is so that others can find where mistakes were made, so they can be grabbed again. If you make digital copies of things and then store them away for 25 years, you may find later that you made some terrible mistakes.

I have cases where I have grabbed a disc—there’s one right now—and somebody told me about the disc that I have grabbed four years ago that it doesn’t work. It works to me, but they’re like: “No, no, no, it’s missing this part and this part.” So, I have to go back to the bins, find it, and use a different method to grab it.

Those kinds of surprises will catch you a lot, which is why I don’t get rid of the originals… In general, turning it into a single digital file that lives on a hard drive works very well, except for where it really does not. And sometimes you don’t get to know until later, which is why I always insist that people keep originals.

There’s a huge difference in how well different types of software are archived. Game software gets a much better treatment—although it’s not enough, which is what’s scary to me; it’s the best, and it doesn’t get enough. Then you get business software like word processors, then you get industrial software, which was written for a very small number of customers… And then, the worst one: custom software, which companies are making and setting up for one or two customers. When that company decides to get rid of it, there’s no clearing house, there’s nowhere it’s going to go to.

Each of these have a problem. Industrial software, which provides access to databases, catalogues or retrieval of historical items, is almost forgotten… Occasionally, I would get a collection from somebody who ran a business, and they’ll have a stack of word processing software, or art or utility software. I will personally image it as fast as I would’ve imaged any game—in fact, faster.

But very rarely I’d get my hands on industrial work, because it usually comes with a contract and an agreement not to distribute. And it makes the person very nervous, because they feel they’re not allowed to and that’s the company’s property.

• • •

The dream we have is being able to put in a DVD reader in the browser and have it play a DVD. It would have menus, it would have the subtitles, it’d have everything! It would have features! What a fantastic idea. We have tried it, I have tried prototyping, and in every single case, you have to take full contents of the disc and re-compress it into a proprietary package. Not proprietary like “secret,” but as its own special version, with a lot it gotten wrong.

And here’s the other thing: DVDs changed their format a lot. You’ll need the power of a VLC [a long-running media player software project which supports many existing formats. — Ed.] or a lot of DVD readers in software to be able to read them because of all of these secret, one-off issues they all had to do their research on. You cannot just write one from scratch anymore: it’s too complicated.

• • •

It’s one thing to know what you are getting. Like, you want to study this famous typewriting program, so you go to it, you find it, you run it, now you’re researching or testing it—you knew that going in. But what do you do when there’s a whole bunch of programs that you don’t know what they are, and you don’t even know where you’re looking?

Part of what we’re trying to do is making sure that the time from thinking that you want to look at something to looking at it is very short. <…> At the Archive, we have a browser that you can click on, and it will show you the contents of a disc as a clickable directory. You can link to the file within a disc image or a ZIP file. Some people use that, some people don’t, but we’re trying.

Sometimes it’s inconsistent, and when you’re dealing in tens of thousands and millions of items, the potential of getting it wrong is huge. I think it’s worth doing. If somebody is coming to an archive like this expecting to be disappointed, they will find plenty to be disappointed. But if someone comes hoping to get better and faster access to some of this history, we’ve done a pretty good job at getting a large amount of that history very accessible.

Interviewed and edited for length and clarity by Yuri Litvinenko.

The Opposite the Editorial section publishes guest columns by authors not affiliated with 30pin’s regular operations. Views expressed in this section are personal and do not have any effect on the 30pin coverage.

The access to any technology, when analyzed on a scale of the world, is never universal. But out of all electronic devices, TVs are most worthy to be described as globally adopted. The common nature of television made it a popular target of comparison for emerging consumer devices, whether by marketing them as no more complex than using a remote control, adopting the broadcasting parlance, or both.

Nintendo’s Wii, a 2006 video game console, checked both boxes from the start. Its novel controller, the Wii Remote, undeniably looked like a cross between a TV remote and an iPod, while the controller’s pointer and motion capabilities introduced electronic entertainment to a plenty of people who wouldn’t otherwise touch video games.

But the extent to which the Wii relied on a television legacy suggests that Nintendo treated the TV better than just a device which popularity they needed to reach. The TV culture was also a source of inspiration, resulting in a user experience which would look out of place if not interconnected with the said culture.

“Companion to the Television”

The so-called seventh generation of game consoles, which the Wii was a part of, was the first one which got rid of the “press start” approach to the user interface. Inserting a game media into a console and powering it on, like it was done since the 1970s, could not be assumed anymore to be an integral part of the experience. A player could as well launch a game from the internal storage, use the console to play videos over the Internet, or even let it idle to download updates.

Facing the pressure from media appliances and computers, console manufactures had to turn their game-playing machines into multipurpose devices. User interfaces of their preceding products, as introduced five years prior seventh-generation consoles, were limited to save data managers and settings screens. However, Nintendo’s competitors beat it to the market with new interfaces. Sony’s XrossMediaBar menu system, used in the 2006 PlayStation 3, was first shipped with a DVR-game console combo almost three years before the new console’s launch. As for Microsoft, they found a way to stream their Windows Media Center made-for-TV platform to the original Xbox console from 2001.

To make up the leeway, Nintendo assembled a team led by Takashi Aoyama and Tomoake Kuroume, who had been responsible for video game menus and console startup screens. The team started their research with designing an on-screen keyboard, and then moved on to creating a way to select a venue for Wii’s programs and functions (which Aoyama described as “content outlets” in his panel talk at the Game Developers Conference in 2008).

“We thought if we succeed in creating a machine that people could think of as the companion to the television, this would be a great way to expand the user base. We never really thought of Wii as fighting with the television for consumers attention. Instead, we thought we could make watching television even more enjoyable,” he said.

The idea of calling content outlets “channels,” as well as designing them up after TV screens, was coined after a remark by an unnamed Nintendo employee as the team settled on lining elements up. “Right then, someone casually said, neither as a complement nor as a criticism, ‘That's a TV channel,’ and that’s when it clicked,” said Kuroume in an Iwata Asks interview.

Dissecting the Channels

Channels published for the Wii included tutorial videos, game patches, and—only in Japan—a food delivery service. But out of all Wii channels, News and Forecast Channels highlight their TV lineage the most. In addition to fitting themes of their content (using a TV to check the weather or news makes too much sense), these channels used flashy transitions, chyrons, and other elements characteristic of news-based TV channels.

These channels, as well as other Wii software, are specific in their features to the audiovisual culture of Japanese TV, says Dr. Ryoko Sasamoto, an associate professor in the School of Applied Language and Intercultural Studies, Dublin City University. “In many ways, Japanese culture is very much visually oriented, and the Wii examples... definitely have similarities with the media practice in Japan,” she notes in an email interview with 30pin.

A researcher of telop (open captions used by Japanese broadcasts and variety shows), she made a note of a digital clock found in a top-left corner of Wii’s informational channels. According to Sasamoto, the feature has the same feel as the clock telop used on Japanese TV since 1956. “Japanese people do heavily rely on this time telop,” she said, describing that the clock mark is displayed at busy times of the day: early in the morning, at lunchtime and in the early evening. “It is changing now, but many Japanese families keep their TV when they are home, and having a clock on the screen really helps.”

Everybody Votes Channel, a mass online questionnaire released later in the Wii’s lifespan, was described by Sasamoto as having “the flavour of game shows and some news programmes,” considering its use of charts and the drum roll SFX. “The music is also very ‘NHK’!” she said, referring to Japan’s national broadcaster.

The Wii menu shares the font style with Japanese TV programming, as well as digital and print publications and public signage, says Kaihatsu, a pseudonymous YouTuber behind the Fontendo project which highlights fonts used in Nintendo products. “The reason this style of font seems to be everywhere in Japan is the same reason fonts like Futura, Univers and Helvetica enjoy wide usage here—it's a very modern, easy to read typeface which comes in many weights and looks good both in headings and longer pieces of text,” they note in a 30pin interview over the email.

The particular font used by Nintendo is Rodin NTLG (New Type Labo Gothic) by Fontworks, a foundry which sans serif Gothic fonts rose in popularity due to their “competitive pricing for smaller clients and individuals,” and due to the high quality of their Rodin font, notes Kaihatsu.

In the TV industry, NHK finalised the switch to a Rodin font for their clock telop in October 2007, according to an archived Q&A from their web site. Before that, some broadcasts used the segmented LCD font to display time. “Interestingly, the clock on the Wii menu uses an LCD font which may have been created to mimic the clock on NHK, as the change to Rodin happened after the Wii's release,” noted Kaihatsu. The clock was added to the Wii menu in an August 2007 update.

Keeping an Eye on the Competition

Nintendo’s strategy of capturing the market share, with a focus on new control schemes rather than drastically improved graphical fidelity, was different from the one of its peers. While concurrent consoles from Sony and Microsoft were by no means unsuccessful, the Wii surpassed them in popularity, surpassing more than 87 million units shipped by the sunset of the generation December 2012. That pushed Nintendo’s competitors into adopting some concepts popularized by the Wii—which, to a limited extent, included the channel metaphor.

Part of Sony’s original pitch for the PlayStation 3 was its raw processing power. To drive the point across, Sony collaborated with the Stanford University in creating a console client for its Folding@home distributed computing project. After the cooperation earned helped both parties set a Guinness world record, Sony updated the client with media features and news content, rebadging it as Life with PlayStation in 2008.

Life with PlayStation, just like Wii’s News and Forecast channels, spread out the data on a three-dimensional model of the Earth. Sony’s globe was more photorealistic than the one on the Wii, which earned the app a positive preview coverage. The content, however, was put to a limited use: there were no forecasts, no news slideshow, and the interface had no visual elements aside from the 3D globe. As a content channel, Life with PlayStation was more limited than what Nintendo offered.

Microsoft took a more media-centric approach by licensing existing TV show formats and offering their video game versions under an Xbox Live Primetime initiative. Games based on TV shows were nothing new by that point, but 1 vs. 100, the only game released under the program, was the first one to offer massive multiplayer, live events, and seasons—a combination of concepts popularised later by Fortnite and HQ Trivia. The free-to-play game for Xbox 360 console, based on am elimination quiz show of the same name, ran for two seasons from 2008 to 2009, with live sessions letting 101 players to compete for a stash of Microsoft store credit.

1 vs. 100, the only game ever run under the initiative, made such an impression on the gaming community, Microsoft reps were answering questions on the show’s potential comeback for years. In 2014, Dave McCarthy, General Manager for Lifestyle Entertainment at Microsoft Studios, told Polygon the company might do something like 1 vs. 100 for their then-new Xbox One system. More than six years later, Xbox head Phil Spencer hinted at making another trivia game “like from our past” for yet another next-generation Xbox Series consoles, as reported by VGC. None were released so far.

Cheating On the Silver Screen

The cultural partnership between the Wii U, Nintendo’s 2012 home video game console, was different than the one established by its bestselling predecessor. The Wii Remote was relegated to the status of a secondary controller, while the GamePad—a big, feature-packed, one-per-console controller—was billed as the Wii U’s defining feature.

The Wii U GamePad had its own 6.2-inch screen, as well as sensors used to evaluate its physical orientation. With game compilations like Nintendo Land or downloadable 360-degree panoramas, Nintendo pushed the GamePad as the means to see beyond the borders of a TV screen. Some games used the GamePad to display supplementary data, like maps and item lists, and most allowed to use the controller as a TV screen doubler. But multi-pane experiments like Star Fox Zero made players who ignored the GamePad screen in a disadvantageous position.

The relationship between the Wii U’s main controller and the TV was adversarial in a way not seen with the Wii or any conventional home game console. And yet, by the virtue of branding, the Wii U had to represent at least some part of the Wii experience. Nintendo, unlike its competitors in the early 2010s, chose to ensure that most commercial Wii games and peripherals could be used with their new console. The Wii U could even reboot into the “Wii Mode,” which presented the Wii menu to the player with minimal changes compared to the 2006 version.

The Wii Mode was used to launch games made for the Wii only, and the interface to run the Wii U software differed drastically from the wall of metaphorical TV sets. Gone was the concept of channels: the software installed to the Wii U was displayed as a grid of non-interactive icons. By default, the icon grid was moved away from the TV to the GamePad’s touch screen, turning it into a tablet as far as a basic selection experience goes.

Some principles which Wii developers swore upon were not followed through with the Wii U. Kuroume once called it “unacceptable” for Wii to display a user select screen upon startup. “If we chose to, we could allow each family member to have their own account. But that just isn’t what Wii is about,” he said in an Iwata Asks interview, unaware that the Wii U would do exactly that.

Down With the TV, Hooray for the Content

There was one thing that was aimed at maintaining the link between a TV and a new Nintendo console: the infrared port on the Wii U GamePad. Nintendo did experiment with ways to control a TV from the Wii, but the TV remote functionality of the GamePad marked the first time when said feature was shipped as an integral part of the experience. Since the GamePad had some processing power on its own, it could be used as a TV remote even when the console was off.

In the US, Canada and Japan, the IR blaster functionality was complimented by online TV listings provided by the Nintendo TVii app. The Western version of the free service provided a universal search across TV channels and streaming services, a sports tracker, and some social functions. Even though TVii was missing some features on its launch, the press coverage of the service was positive: Consumer Reports called it “terrific” while AP’s Ryan Nakashima described it as “an advance from the mass of buttons.” However, that didn’t prevent Nintendo from discontinuing TVii in North America in 2015, never offering it in Europe, and phasing out the Wii U in its entirety due to poor reception by 2017.

Focusing on TV content rather than elements of its presentation was in line with changes in TV usage. Upon the original Wii release in 2006, Netflix was still a DVD rental service, Google had just finalized the YouTube purchase, and TV sets which could stream local media guaranteed a hefty premium. Six years after, TVs themselves—just like game consoles before them—have started to assume functions of other computerized devices. Initial versions of smart TV platforms showered tube watchers with noisy, widget-laden user interfaces stripped off the traditional television decorum. For a moment, the wider visual entertainment industry seemed to be concerned more with collecting as many outlets as possible rather than with its presentation, just like the Wii U makers were concerned with encompassing every use case and control scheme possible.

On its first major attempt at making a general user interface, Nintendo captured the zeitgeist of television, both domestic and international, as the whole way of interacting with TVs was about to change. Its successor, in contrast, looked as if being pulled in different directions. But wasn’t “pulled in different directions” a spirit of early 2010s tech and media industries?

Few things are as disheartening to a community of buyers as an announcement of discontinuation, even more so if the community is passionate and the decision is announced on Friday. That’s what happened with Logitech and Harmony universal remote users last week, as the peripheral company said that it has stopped designing new Harmony remotes and won’t manufacture them in the future.

A programmable universal remote—the one used to control a tandem of devices at the same time, rather than, say, replace a lost or damaged one—is a product for the specialist market of home theater owners and A/V enthusiasts. Wirecutter, a recommendation outlet owned by the New York Times, vouched for Harmony remotes in all price ranges—and had to “return to the drawing board and start fresh” once they become aware of signs of discontinuation. (The editor’s note had since been updated with Logitech’s statement.)

Logitech did say they are going to provide access to software and services, as well as update the device database, “as long as customers are using it.” But with customers reporting cancelled orders and depleting retailer storages, it’s safe to mark the end of Harmony’s 20-year history—and briefly review it.

The first Harmony remote was made in 2001 by Eazy Zapper, a Canadian startup later renamed to Intrigue Technologies. The company was founded in 1999 by Glen Harris and Justin Henry who, according to archived press releases, were frustrated by the number of remotes required to operate a home theater. All Harmony remotes since the original one allowed to select an “activity” (for example, “watch a DVD” or “listen to music”) which allowed sending actions to several devices in succession. To program an activity, a Harmony owner had to enter their A/V devices into the remote using a computer.

Eazy Zapper was not alone in making computerized universal remotes. Philips, a long-standing party in a home electronics business and a de facto standard setter, introduced its first Pronto remote in 1999. The Harmony wasn’t even the first programmable remote with an activity function: Harman, in cooperation with Microsoft, introduced the Take Control remote in 1998, releasing it under Harman Kardon and JBL brands. Both remotes had a touch screen with the ability to rearrange virtual buttons as one sees fit, and, unlike the Harmony, were described by the Washington Post as requiring no out-of-box setup for initial usage.

What set the Harmony apart, according to early 2000s reviews, is its ease of use, highlighted by the original model’s straightforward design. The remote’s ability to remember the state of controlled devices, patented and marketed as a “Smart State Technology,” was another unique feature of Harmony clickers. Initial revisions of the remote could also download listings from the Internet and save (or, in the Harmony parlance, “zap”) TV programmes to keep a note of what’s playing.

The Internet connection was required not only for Harmony’s added functionality. To set the remote up, its owner had to create a Harmony account and use a web service to enter their home appliances data to the remote. Incidentally, the site went down in the middle of filming a TechTV review of the remote in 2002. (Skip to the 2:54 mark.) Since then, every Harmony remote uses either the web or an app to set the configuration up, which makes the 2021 discontinuation announcement especially worrying.

But in 2004, there were no signs of demise. Intrigue was bought by Logitech for $29 million in cash and, per Logitech’s press release, a “possible performance-based payment.” Under Logitech, Harmony remotes underwent several design changes. A small LCD screen, unconventionally placed on the bottom of early Harmony devices, was moved to the top. The zapping functionality was eventually discarded, and the key layout became populated with numeric pads, shortcuts, and—in one case—a set of Xbox face buttons. Two Harmony models, 1000 and 1100, went as far away from the initial concept as to copy the look of premium Pronto remotes, making the device look like a mid-2000s portable video player.

Logitech’s latest universal remotes are different from ones released 10 or 15 years before, but all of them are equally contemporary to media and handheld devices of their time. The original Harmony remote was compared to a Nokia handset, while the 676 model, released in 2004, brought the idea of snappable front panels from the cellphone world. And while the modern, Alexa-enabled Harmony Express was a minimalist device in line with late 2010s design trends, the Harmony One screamed 2008 with its piano black finish and faux-3D on-screen buttons.

Since early 2010s, standalone universal remotes had to compete with devices they were supposed to control. Infrared blasters appeared on TVs by LG and Samsung, as well as the Xbox One console. In 2016, Samsung replaced the blaster with the One Remote, a minimalist clicker which could be configured to control devices connected to the TV. Devices with an HDMI connection to the TV could be controlled by the television and vice versa, provided they supported the CEC protocol. (Sony’s PlayStation 5 console got the HDMI-CEC functionality with a software update the day before this publication.)

With some Harmony devices released within the decade, Logitech leaned even more on the Internet connectivity than before. That made the Harmony Link, a 2011 app-controlled infrared blaster, vulnerable to the encryption certificate rot. As the Harmony division, per its head Rory Dooley as quoted by Ars Technica, “refocused development resources on newer technologies,” the device stopped working in 2018. And last year, Logitech disabled the Harmony Express 17 months after its introduction.

“I think over time, you’ll have fewer and fewer people who feel like they really need that universal remote,” said Logitech CEO Bracken Darrell in 2019 in a discussion with The Verge. He referred to the universal remote division as a “small business,” comparing the size of Harmony business to about 6% of Logitech’s keyboard one.

Without any additional purchases, a modern connected home has a lot of ways to control A/V and home equipment beyond the remote: voice assistants, smart home infrastructures by Apple and Google, as well as the variety of vendor-specific apps. With most Harmony devices, what was left of their once-unique offerings were basic ergonomics of a grippy, multi-key remote. To some companies, these features alone were important enough to bear the costs of replacing Apple’s controversial touch-enabled Siri Remote. To Logitech, whatever drove people to their Harmony remotes by 2021 was not enough to keep the R&D going.

If we chronicled every move Microsoft has made to capture markets beyond general computing, the record would span 30 years of one-offs, obscurities, and commercial flops. While the company did secure its position in the gaming market with the Xbox, its other home appliances and personal gadgets were far less successful.

Microsoft can’t be blamed for not trying hard enough, though—at least, if we are talking about the sheer variety of approaches. They tried both following the leader and coming up with novel concepts; collaborating with household brands or dismissing them aggressively; acquiring companies or licensing technologies to them.

One of the things Microsoft had to offer to gadget makers was an operating system which, at a glance, closely resembled their most famous creation: Windows. In the late 1990s, the four-color flag, already synonymous with PCs by that point, appeared on the predecessors of modern phones and tablets. That system continued to be in use decades after its first release, but with none the public recognition Microsoft was once pushing for. What happened?

“Why Are We Doing This?”

If memories of going shopping for groceries aren’t completely forgotten in this lockdown-induced haze, you might recall the industrial-looking interface of a cashier’s terminal, or a barcode scanner mounted in the aisles. And at many points on their way to the store—or your home, if we are talking about deliveries—those products were scanned using a clunky, sometimes gun-shaped mobile computer.

Chances are high that all of these devices, no matter if they were made five or fifteen years ago, are running Windows CE, a low-profile system licensed to embedded hardware manufacturers. It was developed in parallel to the desktop Windows, rather than being derived from it as the name suggests.

Since the mid-1990s, PC manufacturers licensing Windows have been barred from modifying its interface or hiding defining features like the Start menu. In contrast, Windows CE licensors are allowed to pick and choose what they need from the system and put their own user interface over it. That renders the system featureless by design.

This approach differs from the way Microsoft treated Windows CE in its early years. Most Windows CE devices had the system’s logo “emblazoned right on the plastic case,” Fortune reported in 1998, noting that “Microsoft wants consumers and businesses to think of Windows as ubiquitous—not just a figment of the PC screen.” While the system could be licensed by hardware companies only, Microsoft was running magazine ads for it, promising that Windows CE-powered handheld computers were like “having the essence of your PC at your fingertips.”

To push the adoption of Windows CE, Microsoft brought their main asset into action: a Windows 95 user interface, fresh off a reported $300 million marketing campaign. While Windows CE did not sport an exact copy of the desktop’s interface, there were enough elements of it, from icons and taskbar buttons to the bezel-rich nature of each interface element, to evoke familiarity. With the initial batch of Windows CE devices, the effect was highlighted by the hardware design: so-called “Handheld PCs” were designed to look like miniaturized copies of contemporary notebook computers.

The idea of an ultraportable computer aiming for continuity with a desktop was novel. Every other mobile computing platform of its time, from Apple’s intelligent notebook to General Magic’s collection of “virtual rooms,” was rejecting associations with desktop computing as much as they could. The idea behind Windows CE was to “prevent users from having to learn a completely new paradigm,” explained Sarah Zuberec, Microsoft’s User Research Manager, in Eric Bergman’s Information Appliances and Beyond. (Zuberec was researching and overseeing Microsoft’s mobile platforms from 1994 to 2005.)

Windows CE was Microsoft’s third attempt to make a dedicated mobile computing platform, and the first one which reached the market. Before that, the company had been simultaneously working on two other projects, codenamed WinPad and Pulsar. Both concepts were previewed in 1993 by Bill Gates, Microsoft’s co-founder and then-chairman, with the work on the former beginning a year prior.

The WinPad was a part of the “Microsoft At Work” architecture which aimed to provide a unified way for office equipment to talk to each other. The main focus of the platform was to let PCs send documents to printers and faxes and get data back from them—say, a real-time printout status. But the wider concept also included portable devices and networked landline phones. “The idea is wherever there is a screen… you are to be able to call up the directory and see who’s who and get the information about them,” young Gates is seen promising in archival footage published by Microsoft’s Channel 9. (Microsoft shut down the blog after this article’s publication, but the video file was still left on their servers.)

“Microsoft At Work for Handhelds,” which WinPad was called in the media, was based on the same instruction set as Windows 3.1, a 16-bit operating environment and a predecessor of modern Windows versions. Its later editions, meant for software developers who tested prospective platforms, were found alongside pre-release versions of 32-bit Windows 95. Both systems were designed for ordinary PCs, and the WinPad device had to be a PC as well—not a trivial task at the time when notebooks were just starting to be reasonably-sized. The software project was built around a chip set made by Intel and VLSI, codenamed “Polar,” which was supposed to replicate the desktop 80386 processor in a low-power package. The joint hardware project was canned before any WinPad device came close to being shipped.

While WinPad and the At Work initiative were created for Microsoft’s business clientele, the Pulsar project was meant to have mass appeal. Gates publicized the so-called “wallet PC” concept, describing a small portable computer which would replace IDs and credit cards with messaging and organizer capabilities, in 1993. Nathan Myhrvold, founder of Microsoft’s Research division had pitched him the concept in 1991. “There is no reason why you can’t just carry a single little PC about the size of a wallet—a little flat-screen device that has all your authentication credentials,” as Gates explained in his speech at the University of Washington (as quoted in The Seattle Times).

Pulsar was the first attempt at implementing Gates’ vision of a wallet PC, said Harel Kodesh, Microsoft’s former VP of Consumer Appliances, in an email interview with 30pin. “It didn’t look like Windows; it didn’t share the same API as [the 16-bit] Windows… The team had a feeling something big was happening.”

Gates described the wallet PC numerous times, referring to it as a vision of the not-so-distant future rather than a project in development. Gates’ 1995 bestseller The Road Ahead (which Myhrvold co-authored) laid out one of the most detailed descriptions of the wallet PC. Today, this account reads like an eerily accurate description of a modern phone, made long before even basic cell phones were commonplace. “Your wallet will link into a store’s computer to allow money to be transferred without any physical exchange at a cash register,” said Gates in a passage which would fit modern tap-to-pay systems like Apple Pay. At a later point, he imagines a wallet PC user asking the device for directions to the nearest Chinese restaurant, just like one would ask Google today.

Microsoft was not working on Pulsar by the time The Road Ahead hit store shelves. WinPad was no longer in development either, as both teams were reorganized in 1994. In John Murray’s Inside Microsoft Windows CE, Robert O’Hara, a WinPad development lead, said Gates was disappointed with the lack of shared direction. “I’ve got WinPad and Pulsar working on handheld computing, doing two completely different things. Why are we doing that?” he said, quoting Gates.

Brad Silverberg, Microsoft’s Senior VP of Personal Systems, was tasked with building a single team. “Neither [project] to me seemed like it was on a path to success, though each had some good ideas,” recalls Silverberg, famous for shipping Windows 95, in an email interview with 30pin. To him, Pulsar was “a long way from being a product.” (Murray’s book quotes more direct criticism of the project, with Silverberg reportedly blaming the Pulsar team for not doing the market research and comparing their creation to a student project.) However, he appointed Kodesh, who oversaw Pulsar, as a leader of the merged team. “I chose Harel and a number of WinPad people left the project, which is understandable in these situations,” Silberberg says.

“Fortunately, the Products Sucked”

The principal hardware design was one of the first things agreed on, as requested by Microsoft’s existing partners. “In 1994, HP and Compaq started to ask us to build a professional device,” Kodesh recalls. The former had already been making their “palmtop PC” mini-notebooks, and the new handheld PCs inherited the same clamshell design. “The form factor was considered a winner among knowledge workers who wanted a keyboard and ability to type fast.”

Before the team settled on a desktop interface, the new system “did not have the affinity necessary to identify itself as a Microsoft Windows product,” wrote Zuberec. User interface prototypes published in Information Appliances and Beyond show the main menu with a grid of icons and a bar of ever-present, silk-screened buttons. (On many touchscreen devices of that era, the touch panel was bigger than the LCD display behind it. Silk-screen printing was used to add buttons and control elements to the extra touch-sensitive space.)

The team originally wanted to build the user interface around the concept of documents. The goal of the prototype, according to Zuberec, was to test whether a user would see it as a container for documents or as an application shortcut. Initial experiments, though, revealed that users were split on how to treat the interface. Zuberec attributed the split to the fact the test group was made up of “early Windows 95 users”—that system, as well as every other Windows after it, supported both paradigms.

According to Inside Microsoft Windows CE, the Windows 95-like user interface, which Zuberec quoted as being adopted after feedback from “top-level marketing and executive staff,” was prototyped over a weekend by Tony Kitowicz, the project’s shell lead. “It was all from scratch, brand new,” the book quotes him. Kitowicz felt it was a compliment when people viewed Windows CE as a port of the desktop system.

The first batch of Windows CE devices was released in the United States in October 1996. By that time, the mobile computing market was upended by Palm, then a subsidiary of U.S. Robotics, and their Pilot “connected organizers.” Palm averted the need for a keyboard by using “Graffiti,” a stroke-by-stroke handwriting recognition system. Instead of trying (and failing, due to 1990s tech limitations) to decipher a person's real handwriting, Graffiti made them learn a simplified alphabet where every letter was written in a single stroke. This not only made the text input reliable, but also allowed Palm to build a compact, candybar sized device with a big touch screen—not dissimilar to modern phones.

Palm was aware of the resources Microsoft was putting into Windows CE. Ed Colligan, former VP of Marketing who later became Palm’s president, was quoted in Andrea Butter’s and David Pogue’s Piloting Palm, “We were thinking, ‘Geez, how are we going to compete with this?’ Fortunately, the products sucked beyond our wildest dreams.”

The contrast between Windows CE’s desktop-like appearance and the limitations of the platform made things worse, argued Conrad Blickenstorfer, co-founder of Pen Computing Magazine, in a personal interview in 2018. (He reaffirmed his stance to 30pin before the publication.) “No windows, no cascading, no resizing, no minimizing, no mouse support… All this only added to the frustration of ‘seeing Windows,’ but really only having a tiny and very slow part of it.”

By the end of the year, only NEC and Casio shipped their Windows CE devices, accounting for a bit more than 3% of a market share combined (according to the Dataquest report quoted in a Harvard Business School case study). Meanwhile, Palm shipped more than 360,000 units, capturing more than a half of the market after making it 2.6 times bigger almost all on its own.

Computers, Consoles, Car Stereos…

The next version of Windows CE, released a year later, added support for color screens, improved synchronization with PCs, and added new “pocket” versions of Microsoft’s desktop apps. It also served as a basis for Microsoft to counter the threat of being outperformed by Palm. In 1998, an amount of hardware form factors on which Windows CE was allowed to run expanded notably. Among them was a “Palm PC” (renamed to “Palm-size PC” after a lawsuit from 3Com, Palm’s new parent company), a hardware design with neither keyboard nor a clamshell case.

The demand for a Palm-like device, according to Kodesh, was expressed by original equipment manufacturers, hardware companies working with Microsoft. “OEMs started to ask us for a portrait, smaller form factor, probably because of Palm. And they thought, as well as we, that relationship to Windows would be an asset, so we continued what we did with H/PC.” The team tweaked the Windows CE interface to fit the smaller screen but preserved the resemblance to a desktop system. Even the taskbar was kept, although, at 240 pixels wide, it barely fit the Start button and a notification area.

The first time Microsoft moved away from mimicking ordinary Windows was when they had to put Windows CE into a car. The “Auto PC,” unveiled the same day as Palm PC, was aimed at combining the best of late 1990s in-vehicle infotainment, such as CD playback and turn-by-turn navigation, with voice commands, hands-free calling, and connectivity with Windows CE handhelds. Most powerful Auto PCs, such as the $3,000 Clarion Joyride, used desktop-class processors to display maps and play DVDs and MP3s.

“In the car, driving is the primary task. Any interaction with the Auto PC while driving is considered a secondary task, and input and output methods were designed with this in mind,” wrote Zuberec. While top-of-the-line systems had their own large screens, the base hardware (like the first Clarion AutoPC, with a suggested retail price of $1,299) had to rely on a screen smaller than a single-DIN faceplate. “Viewing angle and distance defined the font size. Hardware that fit into the standard car stereo slot allowed very few hardware controls.” And yet, some Windows iconography was preserved, including a hardware Start button.

A less obscure example of putting a Windows CE logo on a device outside the realm of computing was the Dreamcast, Sega’s last video game console. Under a deal with Sega, Microsoft adopted Windows CE for the console while Sega declared the compatibility on the front of its case. The collaboration was meant to allow developers to “create cross-platform titles more efficiently” by using the same tools and instruction sets to make games for both Windows and Dreamcast. It wasn’t quite the same promise Microsoft would make later with its own Xbox, but the direction was similar.

Microsoft’s instruction sets and libraries, used in Windows CE for Dreamcast, were intended to ease the cross-platform game development. Developers are still using similar “middleware,” albeit without a full-fledged operating system tacked on, with its licensing often requiring developers to disclose the use of the technology in a game’s splash screen or credits. But the terms of the agreement between Sega and Microsoft yielded the latter a level of publicity unprecedented even today. The Windows CE logo was not only put on Dreamcast’s case, but within a startup screen for every game which relied on it.

Most games didn’t. Of 600-plus games officially made for Dreamcast, only 78 used Microsoft’s game development option. The others were made using Sega’s homegrown tools. Tom Clancy’s Rainbow Six, ported to the Dreamcast in May 2000, missed the console’s launch because of a “lack of experience” by the porting team and “issues surrounding the use of Windows CE,” CNET reported. Besides, the Dreamcast itself did not have Windows CE built in: it had to be kept on and loaded from the same one-gigabyte optical disc as the game itself. Developers had to account for this overhead in storage and performance, too, as Windows CE itself was using a bit of console’s resources.

“Dilutive to Microsoft”

Shortly after its launch, Windows CE went from supporting a single, strictly defined type of device to powering a whole range of devices, mostly incompatible with each other on both software and hardware levels. That range, put into a single picture, would’ve impressed any manager focused on scale and corporate unity the same way as Google’s homogeneous corporate-colored app icons would today. But within Microsoft, there was an attempt to rethink the expansive strategy and revert Windows CE to its original form.

The 2001 book Breaking Windows by David Bank describes the “Starting from Scratch” memo, sent by Kodesh to Bill Gates and Steve Ballmer. The Windows CE lead proposed to leave the original system for use with handheld PCs only, kill off its other flavors, and then create a division within Microsoft which would have worked with hardware vendors to create a tailor-made embedded software—namely, the one for cell phones. Gates rejected the idea.

“I was losing the war there,” tells Kodesh 30pin. “We have not moved enough from Windows—that was both an asset and the liability.” He argues that people like the Windows experience because they know how to operate within it, but it didn’t allow Microsoft’s mobile team to simplify it further. According to Kodesh, if he wouldn’t leave the company in 2000, his next battle would have been “diverging from the Windows interface.”

Before leaving the company, Kodesh shipped a revamped version of a palm-size PC platform. Dubbed “Pocket PC,” the new software was stripped off several desktop interface concepts which didn’t make sense either on small screens (like the taskbar) or, in retrospect, on any device which would fit into a pocket (double-tap to activate icons). The system’s appearance was streamlined as well, with bezels gone and toolbar icons flattened. The system kept being visually inspired by Windows, but the sense of squeezing as many desktop elements as possible into a 3.5-inch screen was not as pronounced as before.

The release of the Pocket PC in April 2000 also marked the moment Microsoft started to move the Windows CE brand away from the consumer spotlight. The platform, as well as its Handheld PC 2000 relative and a later Smartphone 2002 spin-off, continued to be based on Windows CE. But in the software branding “Windows” was now a descriptor, not a centerpiece, as the device were released with a mouthful of a full name: the Microsoft Windows Powered Pocket PC.

The sidelining of the Windows CE brand was explained to the developers a week after the launch of the Pocket PC. It was a part of a wider reorganizing within Microsoft, with divisions responsible for the Windows CE and embedded versions of the desktop Windows combined under the helm of Bill Veghte. A new brand, “Windows Powered,” was meant, per Microsoft’s press release, to “simplify and unify all the Microsoft embedded and appliance solutions.”

Microsoft’s executives “were convinced that Windows should grow up, not down,” said Kodesh, recalling the pressure from Microsoft’s corporate marketing. “There was a concern that the Windows CE is dilutive to Microsoft.” He points at the feedback from salespeople who “started to complain they are trying to sell laptops, but people want the lighter, thinner, instant-on devices”—just like notebook-sized “Handheld PCs Pro” made by Windows CE licensors.

“Windows CE never had a chance because Microsoft made sure to keep it nearly useless, so as to not compete with Windows,” argued Blickenstorfer. To him, Microsoft was never committed to pen-based computing “because all [it] was ever interested in was the desktop and laptop version of Windows.”

The Windows brand wouldn’t return to the company’s mobile efforts in earnest until 2005, when the Pocket PC and the Smartphone platforms were partially unified as “Windows Mobile.” The Windows Powered name remained an obscurity, most prominently seen in custom, enterprise-ready installations of the Windows 2000. As for a new interface which wouldn’t rely on familiar Windows concepts, that took another five years, a new generation of mobile platform, and the release of the Windows Phone 7.

“That Thing Called CE”

The shift in the way Microsoft handled the Windows CE was seen not only in marketing. After the Handheld PC 2000, the company stopped providing hardware manufacturers with the original version of the system. The Auto PC platform did not get a new standard user interface either, with later automotive applications of the Windows CE being mostly stripped of their Microsoft identity. (Some versions of the Blue&Me infotainment system, used in cars by pre-merger Fiat Chrysler Automobiles well into 2010s, had a Windows key on a steering wheel, though.) Microsoft did make a new Windows CE platform for portable media players, but it was discontinued after Microsoft decided to compete with Apple’s iPod on its own.

“Microsoft were always chasing what made the most commercial sense. They experimented with different form factors and ultimately chose the same path as everyone else,” writes Chris Tilley, editor-in-chief of HPC:Factor, a site that has served as a community hub for handheld PC users for the last 20 years. In an email to 30pin, he argues that Microsoft should have followed with focusing on the clamshell design. “The H/PC was—and in fact to many still is—a viable productivity tool,” he notes, describing tiny Windows CE-powered notebooks as “creation devices” as opposed to “consumption devices.”

Tilley thinks the era of Windows CE 2.0 represented the age of experimentation at Microsoft. “No one knew what ‘the’ form factor was going to be,” he said speaking of a flurry of hardware variations supported by the system in the late 1990s. However, Tilley notes that said variety, combined with constant changes in the core system and the high cost of development tools, “fractured the developer ecosystem.” “Microsoft had this thing called ‘CE’ but didn’t really know what to do with it,” Tilley says. “They came up with the technology and then attempted to find a use for it.”

The year 2012 marked the end of Windows CE as a system Microsoft itself would want to build consumer products with. The Windows Phone 8 got the same kernel as the desktop Windows. The move brought the system closer to the vision unveiled alongside the first edition of Windows 10 in 2014: the one of a single system, used on everything from phones and PCs to consoles and Internet of Things devices.

With Windows 10, Microsoft offered developers the way to write a program which, if they would like to, would run on all types of devices supported by the system. The vision lost some of its charm after the phone version of Windows 10 was abandoned after years of irrelevance. However, Microsoft was successful in porting the desktop version to ARM processors, foreshadowing Apple’s move to its own chips with the same architecture. The company has also released a stripped-down version of Windows 10 for embedded devices, complete with a migration plan for clients who still hold onto Windows CE.

In the late 1990s, Microsoft was bringing its secondary system to various hardware platforms; by the 2020s, its flagship software product received the same treatment. But with the upcoming Windows 10X, there is a new set of risks for the company. The streamlined edition of Windows was supposed to run exclusively on dual-screen devices but was later confirmed to power single-screen PCs as well. Alongside with differences in programming instructions and the app support, there is a chance Microsoft would once again create a similar-yet-different system.

With Windows 10X and the future of Windows as a whole, Microsoft faces a different challenge than 25 years prior. But as a precaution, it’s better for everyone to look at the system which ended up in a less prominent position that was once envisioned.

Edited by Samantha Lomb. Special thanks to Ernie Smith for getting a copy of Inside Microsoft Windows CE on behalf of 30pin.

Correction, February 10, 2021: Due to a technical error, a paragraph in the closing section of the article was published alongside its draft version. We’ve removed the latter.

Update, February 1, 2022: A dead link to the Channel 9 video was fixed.

In 2009, a Polaroid Sun 600 camera arrived at a college bar in Ann Arbor, Michigan. The camera, age 28, was older than both the bar itself and the person carrying it. It was under the arm of amateur photographer Marita Murphy, who was grabbing some graduation-week drinks with her younger brother and his college friends. Once inside, Murphy set the camera on top of a round table where her brother’s friends could admire it.

Eventually, Murphy elevated the camera and turned it on some of the curious faces at the table. Someone gasped and said, “Wait, you’re actually going to use it now?” And with that, she triggered the shutter, and the camera belched a loud sna’errrrupp. The Polaroid ejected a single, white-bordered photo containing a candid image of the celebrating graduates. It would take a few moments to develop.

More than a decade later, Murphy, 37, works in radio production in San Francisco. As she shared her story with 30pin during an email interview, she recalled that at that moment, the entire bar turned to watch. “It was really the iconic sound of the Polaroid that stopped people mid-conversation and had lots of heads turning,” she remembers.

Murphy was among the last of those able to get their hands on the original film for this vintage camera. Early in 2008, the Polaroid Corporation—or what was left of it—announced it was going to stop producing film. This wasn’t quite accurate, though. It had already stopped.

People’s passion for instant cameras outlived Polaroid’s intention to make film cartridges. But without them, every one of these old cameras would end up in the landfill. Was there any way to bring the camera supplies back?

A Few Millennial Missteps

The Polaroid Corportation’s downhill slide started many years before Murphy inherited the camera. In 1997, its stock had been trading at $60 per share, but by October 10, 2001, it was trading at just 28 cents. The company prepared to file for bankruptcy protection. What happened?

Polaroid had plenty of time to see that digital technology was coming down the pipeline, and in some ways, it acted on this foresight. According to a 2009 article in Yale Insights, a publication from the Yale School of Management, Polaroid was investing heavily in digital photography by the end of the 1980s. A decade later, its digital cameras were among the top of their class in sales.

But Polaroid made several missteps.

First, according to Yale Insights, the company overestimated consumers’ need for a tangible, printed item. Polaroid’s CEO Mac Booth made this clear in a 1985 letter to stockholders. “As electronic imaging becomes more prevalent,” Booth wrote, “there remains a basic human need for a permanent visual record.”

Polaroid echoed this assumption that printing would remain important fifteen years later in a 2000 New York Times article describing this printer-focused strategy. Then-CEO Gary DiCamillo told the Times, “The cameras are just another way of capturing images, many of which we hope people will want to print.” He sounded a bit less certain than Booth had.

Second, a number of Polaroid’s products released around the turn of the millenium either flopped or failed to come to market quickly enough. These included inexpensive kids’ cameras decorated with Barbie and with the Tasmanian Devil; its pocket camera, which took pictures on a self-adhesive strip; its smaller-sized Joycam, which used old film from its failed SLR camera; and its single-use, disposable camera.

Third, Polaroid’s company culture exhibited a distaste for electronics and, later, for anything digital. It was married to its instant film technology, and it was not interested in adding electronic components to its camera technology. This meant that for Polaroid, chemistry was king. Former Polaroid executive Hugh MacKenzie told Yale Insights, "The culture of the leadership was chemistry and [film] media first.”

It was this chemistry that defined the company. DiCamillo told Yale Insights, “Instant film was the core of the financial model of this company. It drove all the economics—not instant cameras and not hardware or any other product; it was instant film.” The film, DiCamillo noted, had a 65 percent gross-profit margin.

Polaroid’s film technology was, indeed, a marvel of chemistry and design. What we call instant film—those thick, white-edged rectangles we load into the camera—is perhaps better described as a film unit. To produce an instant photo, a film unit that includes a film negative, a positive print medium, and, in the middle of this sandwich, a “pod” containing a chemical developer reagent, is rolled out of the camera. The rollers crush the pod and release the developer reagent. This is what allows a photograph to develop instantly, without use of a dark room.

The first version of the instant camera was developed in 1947 by entrepreneur and charismatic Polaroid founder Edwin Land. In 1972, Time Magazine’s cover story was about Polaroid’s new user-friendly SX-70. It was titled, “A Genius and His Magic Camera.” While instant photography was dismissed by some cultural critics as a gimmick, photography magazines such as The Camera and American Photography; artists including Andy Warhol and Ansel Adams; and consumer public welcomed the new product.

Yet despite having invented these magic pods, it didn’t look like the Polaroid Corporation would survive another decade. As digital cameras became more popular, film and film camera sales continued to plunge. Polaroid was delisted from the New York Stock Exchange, and by 2001, retirees were left with a depleted pension fund and now-worthless stock they’d been in many cases forced to buy. It seemed that a digital image, easily shared, edited, and stored, was the way of the future.

Polaroid changed hands several times after 2001. In 2005, Petters Group Worldwide, busted during the Great Recession for being a Ponzi scheme, bought the company and created a 5-year plan for it based on a “forecast” of the rate at which demand for film would taper off. First, the company would produce all the film this forecast predicted it could sell over 5 years. Then, film sales would support a business model based on putting the Polaroid brand name on objects mostly made overseas, such as LED picture frames. Finally, the instant-photography branch of the company would quietly fade away.

The Impression That Never Faded

But that’s not what happened, though, according to David Hale, Polaroid’s Vice President and head of American Sales who worked for the company for three decades until 2008. In an interview in the 2012 documentary Time Zero: The Last Year of Polaroid Film, Hale explains that Polaroid film consumers kept buying film at a steady rate, not a decreasing rate. By 2008, supply stockpiles were running out well ahead of schedule. On top of this, Polaroid’s film-making factories had been long shuttered, and the bulk of its workforce laid off.

Far from this crumbling empire, at the beginning of 2009, Murphy was clicking through eBay listings in an increasingly difficult mission to find affordable film for her new camera. She had been given the camera by a friend who was excited to have found someone who might use it.

The film Murphy had that day at the bar was original polaroid film, expired, and the photos weren’t quite as sharp as she had hoped. Still, she explained, her Polaroid had stood out from the crowd of digital cameras. It had made an impression, and not just on her brother’s friends, but on all the people who turned to watch the Polaroid at work in the bar that day.

But casual interest from a bar crowd wasn’t what precipitated Polaroid’s turning point. Instead, some of the many enthusiasts appalled by the 2008 announcement that film production would cease formed a group called The Impossible Project. It’s mission? To resurrect an abandoned Polaroid factory in the Netherlands and recreate Polaroid film. A website for a 2020 film called The Impossible Project calls Polaroid enthusiasts “the analogue heroes who dare to swim against the digital tide.”

The Impossible Project was aptly named. An instant film unit must contain every chemical needed for image development. According to the Time Zero documentary, some of those chemicals (such as mercury) had been restricted since the pods were first designed. Plus, old manufacturing processes, which had taken a long time to perfect, would be almost impossible to recreate in the resurrected factory.

Still, The Impossible Project pressed on, with the help of 14 former Polaroid employees. In early 2009, a Facebook group named after The Impossible Project was created in support of the endeavor. By later that year, the group boasted a membership of 3,700. (That number is at nearly 13,000 today).

While Polaroid enthusiasts were working to save analog photography, amateur Polaroid photographers continued buying Polaroid cameras. They found working cameras at yard sales and thrift stores, and in the forgotten corners of attics. And once in hand, these photographers would seek out film, creating ample demand.

Senna Flora, 32, a Tucson-based educator and artist, told 30pin during an interview that “it’s about the novelty.” She uses a 600-series camera she found in an outdoor free-box to photograph friends and family. She likes to give the photos away as gifts. The film is a bit expensive for her. She can only buy one 8-pack a few times a year. But, she said, it’s worth it.

Alana Cox, 30, a small business owner in Brooklyn, has been buying film since before the shortage. She told 30pin in a phone interview that she had “never connected with any Polaroid people online,” meaning that she’d never joined a group of Polaroid enthusiasts. However, she has maintained an interest in Polaroid photography since high school, when she started picking up cameras from yard sales.

“I do love the Polaroid die-hards,” she said, but she herself isn’t too particular about the brand-name of the equipment. For her, it’s about the photos. Instant photography isn’t about having just any kind of printed photograph. It’s about creating “a true, physical, ineditable, moment in time,” she said. She also likes not having to work in a dark room or fiddle with Instagram filters for hours. “With a Polaroid, there’s so much chance to it. It’s up to the Polaroid gods how it turns out,” she said.

Imperfection in the Digital World

If, at the turn of the millennium, Polaroid made the mistake of leaning too much on film and chemistry, in 2008, it was this analog, single use film unit and its fans that had the power to save the company.

The Impossible Project successfully produced its first film at the Netherlands factory in 2010, just before it would have run out of start-up funds. It has continued selling instant film and growing as a business since then. While users on Facebook complain that first-generation film from the Impossible Project doesn’t work all that well, newer generations offer higher (if sometimes uneven) quality and good variety.

In 2017, The Impossible Project bought the rights to the Polaroid brand, and changed its name first to Polaroid Originals, and then just Polaroid. It started producing a new generation of cameras, including the Polaroid OneStep and the Polaroid Now, to rival the Fuji Instax. It also sells all sorts of printers, from ones that fit in your pocket to large format printers; accessories, such as apparel, camera bags, and filter lenses; new digital and hybrid cameras; and a variety of vintage Polaroid cameras with film to fit.

Twenty years ago, it seemed that the brewing obsolescence of chemistry and tangibility—the very things Polaroid stood for—would spell the end of instant photography. Yet now, adrift in a digital world, we seem to long for a reason to look away from the screen, to share something one-of-a-kind with friends, to take a break from filtered and edited perfection. Instant film offers just that.

Edited by Yuri Litvinenko.

Correction, September 28, 2020: A previous version of the caption under Cox’s selection of photos mislabeled them as Polaroid shots. They were actually done with a modern Fuji Instax camera.

A modern phone isn’t defined by a single feature. When compared to their predecessors which set off a mobile revolution in the early 2000s, they share plenty of traits once unimaginable for gadgets of its class—from app catalogs and embedded payment cards to an always-online philosophy which makes the word “phone” itself a misnomer. There is, however, the single, most obvious mark of a cell phone of 2010s and beyond: a touch screen interface which makes it easy for users to type, flick, browse, and scroll their way to a despair.

The combination of a finger-friendly touch screen hardware with fitting controls and gestures was the bigger of two bets made by Apple with its first iPhone in 2006. (The other one was building a mobile browser which, per Gizmodo’s Brian Lam, “actually renders everything ‘as it should.’”). It wasn’t until the next generation of iPhones when they got its seminal App Store—and that’s when smartphone companies joined the fight, investing into new platforms such as Android or adopting aging Symbian and Windows Mobile systems to the new reality.

Smartphones, however, were representing only a fraction of the mobile phone market by then, and many people didn’t really feel the need for one. The cost of ownership, considering both the price of handsets and data plans, was one of factors. Notably, switching between dozens of apps and having a constant Internet access wasn’t yet an unspoken requirement. Owning what would be retroactively called “feature phones” was perfectly acceptable in the late-2000s society.

And yet, feature phone manufacturers weren’t able to ignore the demand for the iPhone and similar devices. The popularity of “touch phones,” which technically predated the iPhone, boomed, and said phones formed a category representing a notable part of the mid-decade mobile phone selection. Select touch phones, complete with a rich feature set for their time, were going directly against the iPhone, with one of manufacturers having an audacity to publicly accuse Apple of copying the whole form factor.

iPhone’s Concurrent Couture Competitor

Apple’s second entry into the phone market wasn’t, by any means, the first touch screen phone ever. IBM, a one-time smart phone maker, used it as the key input method all the way back in early 1990s. And while a touch screen was a feature primarily found on smartphones, some odd, less capable phones did use it for features of varying usefulness—from drawing to entering Chinese characters by hand. What did set the iPhone apart from those phones is the way the touch screen was presented—not only as a primary control method, but also the one which did not need styluses, keypads, any other intermediate or auxiliary elements.

While the iPhone, by virtue of many of its advantages, ended up writing itself into the history, it wasn’t the only finger-friendly touch phone around. By January 2007, LG introduced its “display only phone”—first to the committee of the iF Design Award, then to the public in a much-touted partnership with Prada. With a minimal number of buttons and a clear focus on touch interactions, the Prada phone made it too easy to be compared to the yet-anticipated iPhone. The Korean company managed to release it before the iPhone, albeit starting with Europe rather than the U.S. That let some top LG managers made some bold statements towards Apple.

“We consider that Apple copycat Prada phone,” said Woo-Young Kwak, head of LG Mobile Handset R&D Center, in February 2007. He claimed that the U.S. company saw the design as it was unveiled in the iF Design Award. Few days before, when asked if LG plans to sue Apple, Chang Ma, LG’s VP of marketing, laughed: “Should we?”

For all the showboating, it looked like LG was choosing the distribution strategy contrary to the one of Apple. It started the distribution with Europe and Asia, later entering the U.S. market without a carrier partnership. That, in turn, impacted the price: LG Prada cost $849 unlocked while the more accessible iPhone cost as low as $499. And despite the fact the LG phone had some benefits over the Apple one (namely, the video recording ability which wasn’t present on the first iPhone), the latter had more capable hardware and software.

Unless you liked Prada too much, its co-branded phone was a bad deal overall. But the partnership had not only persisted till 2012 but solidified the touch phone class as a viable one, coexisting with the iPhone and modern smartphones—and, occasionally, trying to best them.

From Overcoming to Undercutting

Over time, touch phone makes were finding its customers alongside the cheaper part of a price spectrum—the one smartphones weren’t able to capture for a long time. That doesn’t mean LG Prada was a luxurious one-off: more touch phones than it might seems aimed for the top.

One of them was the Samsung Jet, released in 2009 with a slogan “smarter than a smartphone.” The audacity was justified: its 800 MHz processor was unprecedentedly powerful for a feature phone of its time. (The phone was capable of running Android, as proved by tech enthusiasts later.) Samsung aimed to make the best out of the hardware, as far as its custom TouchWiz interface allowed, enabling it with a WebKit-based browser, a task switcher, and video codecs which did not require converting videos with a PC. For all the effort to make Jet the next thing after the iPhone (multimillion-dollar marketing plan! launch events! holographic shows!), it largely fell into obscurity: cheaper Star and Corby phones were notably more popular.

The name “Samsung Jet” might have been re-purposed for a line of vacuum cleaners later, but the phone did not exist in a vacuum. While Samsung made its bet on the overall processing power, LG focused on high-megapixel camera sensors with its Viewty and Renoir phones. Smaller companies started to make their own touch phones, going for LG and Samsung’s cheaper offerings. With its first phone of that kind, Alcatel decided to make it as small as possible while sprucing it up with an LED navigation pad.

The market was lucrative even for manufacturers which had never made basic phones before. HTC, a prime manufacturer of smartphones outside Nokia’s realm, released the HTC Smart in 2010 and the updated Freestyle a year later. HTC did not write its own software platform from the grounds up—and neither did LG and Samsung. All three companies, as well as some more, were licensing the BREW system from Qualcomm and writing their own user interfaces on top of it. With the Smart, HTC beat both of their competitors at adopting Brew MP, the latest revision of it—and, in a unique fashion, featured the Brew MP brand in all the marketing and user-facing materials down to the PC synchronization tool.

It wasn’t until 2010 when Nokia made its own touch phone. The company first brought touch input to its tired-and-true Series 40 phone software with Nokia Touch and Type, a pair of candybar phones which combined a traditional numeric keypad with a small touch screen. (The only giveaway these Nokias were touch-enabled was the lack of navigation buttons.) A year later, after releasing the C2-06 slider which looked more similar to contemporary touch phones, Nokia introduced its first “full touch” phones in its new Asha lineup.

In 2013, Nokia moved away from using the Series 40 platform for its touch phones. The new platform, introduced with the Asha 501 and released on four more phones, looked like a downscaled yet evolved version of the MeeGo Harmattan, a prospective Linux-based system briefly used before Nokia’s switch to Windows Phone. The interface was designed by Peter Skillman, who previously worked on the MeeGo, and many of the latter system’s cues, including gesture navigations, were found in new Asha phones as well. Most major social media apps, from Facebook and Twitter to Foursquare, had Asha versions, and the system itself got the OneDrive integration a bit later.

The last batch of Asha handsets, as well as Samsung’s REX phones which looked deceptively similar to their concurrent budget Android phones, were far from the premium mark set by LG Prada and the likes. These touch phones, squarely sold in India and emerging markets, were more capable than ever, and yet they paled in comparison with even the most basic smartphones. The latter have reached the price point of $100 and below by then, trumping the whole class of feature phones.

Back to Basics and Beyond

“I think the simple phones did take the wrong step of trying to emulate and compete head to head with the smartphones, not realizing that what made them a unique offering was their simplicity,” tells Joe Hollier, Light co-founder, to 30pin in an email interview. After a crowdfunding campaign in 2018—way past the prime age of touch phones—the company released the Light Phone II, a no-distraction touch phone with a deliberately limited feature set, an e-ink display, and a $350 price tag reminding of decade-old premium handsets.

With a small body and a squarish screen, it almost looks like a reinterpretation of the Asha 500 phone. Hollier insisted the team did not know of that one before). The Light Phone II, however, was built from the different line of thought, as its its creators handpicked what to select from a modern technological stack. So far, the handset can only place calls, send messages and serve as a hotspot and an alarm clock. Even navigation, which Hollier promises in future updates, will be limited to direction alerts. “The Light Phone II is not anti-technology… it’s more about the right tool for the job,” he explains the rationale behind the decision.

Disregarding the countless no-brand handsets of dubious origin, the Light Phone II is one of few feature phones on the market by 2020s and the only one with a full touch screen interface. The other side of the market, represented by Nokia’s nostalgic remakes and ultra-cheap handsets by an Indian telecomms juggernaut, is powered by KaiOS, a mobile platform specifically designed for keypad-driven devices. None of current KaiOS phones use a touch screen, and a KaiOS Technologies representative stressed in an email to 30pin that “the user interface used in KaiOS today has been designed from the ground up to be optimized for non-touch feature phone hardware.”

KaiOS developers were taking their inspiration from both feature phone and smartphone system, wrote the developer’s rep in an email. The main focus for the project was to to provide smartphone-level apps and Internet services on affordable feature phone hardware; said goal is still the main one today.

A $350 gadget does not exactly compete on the same field as the ultra-cheap phone which might be one’s first Internet-enabled gadget ever. The Light co-founder agrees with that, noting the Light Phone II goes for the quality of its design, not the low price point. With all the differences, he says the Light team has a lot of respect for KaiOS developers: “[We] think they are filling an important need in trying to help make smartphones more available to a larger audience.”

To Hollier, however, squeezing in Facebook and Twitter in a keypad phone “feels like a gimmick,” as these features pushed smartphone users into seeking simpler phone alternatives. That’s where, he notes, previous touch phones missed their chance, as they ended up being “not-as-great versions of those smartphones they aspired to be.”

Could the promise of a no-distraction device revive the touch phone? So far, the Light Phone II is the only phone of its kind, with no direct and affordable alternative in sight. But the current KaiOS-powered feature phones are slowly becoming more inaccessible to younger people, which might have never interacted with a keypad phone. If the goal of bringing the modern connectivity to the next billion of users persists, feature phones will have to adapt to evade another extinction.

This article was first published in January 2019 on Yuri Litvinenko’s personal blog. It has been expanded and further edited to better fit the 30pin style.

Were personal computers ever personal, in a literal sense? HP was once so sure they were, they put hundreds of millions of dollars into their worldwide ad campaign, with a slogan “The Computer is Personal Again.” As adverts rolled by, actually personal computing ended up on phones and across the Internet while smart TVs and consoles took the household entertainment to themselves. The shift eventually made people disregard PCs in general, as there weren’t many reasons to care about them outside of work. After all, it took a forced global switch to working from home to make Windows 10 usage jump 75% year-on-year.

PCs (in an “IBM PC compatible” sense) essentially made a loop back to the 1980s, when they were only interesting to companies and hobbyists. That contrasts with the way computers were sold just a few years later, as a multi-purpose machine for the entire family, the one which can be used for work as well as leisure. The surge in media capabilities coincided with a set-top box boom of late 1990s—and, since computers could now do so much, wouldn’t it make sense to make them a prime home appliance?

And thus, the idea of a “home theater PC” entered the collective mindset of users and manufacturers. They were making computers fitting a media rack (in both style and physical dimensions), coming up with ways to control them from a couch, and generally make them a little bit friendlier than their desktop counterparts. The idea was to replace all specialized media appliances in a household with a PC-based “media center.”

Was the end goal ever reached? Not really, aside from hobbyist projects and select showcase models. And yet, these PCs represented a prominent image home computing was trying to differentiate itself with.

Paint It Black (And Sell It for $4,699)

By the year 2000, the computing industry has been toying with the “media PC” concept for quite a while. CD-ROMs allowed distributing digitized audio, video, and photos, which led to the popularity of the concept of “multimedia.” The selling point of “multimedia” was the combination of several types of content—for the most part, text with sounds, pictures, or low-quality digital video.

At that point, multimedia wasn’t going to disrupt any established audiovisual industries. (Multimedia CD companies were chasing the clientele of paper encyclopedia publishers instead.) Still, humble attempts at bringing computers closer to the media rack were made—notably, by Apple rather than PC manufacturers. In 1993, Apple released the Macintosh TV, a modification of one of its all-in-one Performa models which, thanks to a built-in tuner, could also replace as a TV set. For that purpose alone, Apple made the make the Macintosh TV in black—the main (if only) color of contemporary A/V equipment, and a bold, unique color for any computer in the 1990s. The TV-enabled Mac failed, but the trend of making multimedia computers black persisted for quite a while.

The Destination, presented by Gateway 2000 in 1996, was black as well, complete with a gold logo which mirrored the color combination of many concurrent pieces of home equipment. Instead of an all-in-one form factor, Gateway chose to ship the Destination with a 31-inch monitor, a huge living room display for its time. The Windows 95 PC, which cost up to $4,700 per the New York Times, packed a beefy 120 MHz processor and came with a Field Mouse, an input device which combined a TV remote with a trackball. (A predecessor of Apple Siri Remote if you must.)

While Gateway 2000 was trying to take over all the home media equipment at once (while being sold as the presentation equipment to schools and businesses as well), some devices by other PC manufacturers were a bit more focused. Namely, the iPAQ Music Center and HP de100c were taking advantage of PCs’ newfound ability to rip music CDs and compress them for storage. Expensive yet limited in their functionality, neither them nor their direct successors gained any foothold.

Microsoft Gears Up, PCs Get Prettier

The early 2000s marked the mass adoption of DVD, in part thanks to Sony’s PlayStation 2. Before Apple entered the scene with the iTunes Store, Sony was in an unique position as a conglomerate of tech and media companies. It comes as no surprise they pushed digital media across plenty of their product lines, including PCs. And, since they were still interested in pushing the MiniDisc format, some of their VAIO computers had both DVD and MiniDisc drives and had some neat looks to boot.

Vertically integrated companies weren’t the only ones interested in building a media PC. Barebone PC configurations, made mostly by Taiwanese manufacturers, let people assemble their own computer jukeboxes. Barebones included everything but RAM, a hard drive, and, in some cases, a processor in a partially assembled platform. For DIY enthusiasts, they represented a chance to build a compact, good-looking PC and tune it to their needs.

One of first configurations of its kind was MSI’s MegaPC. Unlike more utilitarian barebone systems made by Intel in 2010s. the MegaPC and its successors had a front-panel display, could play music with actual PC turned off, and were successful in shaking off the beige-case legacy.

Barebone computers—most notably, Intel NUC models—are still relatively popular, and those who want to build a computer from scratch can now choose from a variety of compact cases and motherboards. Amusingly, it almost looks like a PC styled after Sony Ericsson Walkman phones, despite those phones not being introduced two years later.

While PC building community was excited about attractive hardware, Microsoft had its own plans on the software side of things. Released in 2002 and updated over the course of three years, Windows XP Media Center Edition was a modified desktop OS with several new applications bundled, The key one was Windows Media Center, a media hub designed specifically for a TV screen. While being heavily marketed, Windows XP MCE was only available to select few licensed manufacturers. The list of companies and countries expanded release after release, but still, few official Windows media PCs were built. As for power users, they were left relying on third-party or homegrown solutions.

At the same time, Microsoft was also craving for tighter control over the specifications of media PCs. The company set the specification for a Windows media remote, complete with a green, bubbly “flag” button which would invoke the signature ten-feet UI. Microsoft introduced some first-party media PC hardware as well, such as remotes and a media keyboard. Microsoft has also pushed for creating “Media Center Extenders” for streaming videos from a Windows XP MCE machine. These extenders were made as TV set-top boxes as well as software solutions for the original Xbox and Xbox 360.

The next version of the Media Center, released alongside Windows Vista in 2006, was available directly to consumers as a part of some retail versions. That version, as well as the one bundled with Windows 7, was the first made-for-TV PC shell for many people—especially as notebooks became prevalent and HDMI made connecting them to TVs a breeze.

With notebooks, Microsoft attempted to introduce way more unconventional software media standard in addition to the Media Center. Introduced in Windows Vista, the SideShow was a standard for low-resolution auxiliary displays embedded into computer cases. With notebooks, the idea was to let users interact with a computer as its lid was closed, either by reading notifications on a two-line screen or using the larger screen as a media player.

Developers have envisioned other use cases for Windows SideShow, including connected photo frames, detachable media players, and mobile phone sync. None of them, for some reason, involved front panels already used by barebone PCs.  what was already done by barebone PC manufacturers.

Home Entertainment Shifts Away From Computers

As the PC world waged a war on home appliances, Apple (which, in 2004, was making computers on a PowerPC silicon) introduced Front Row, its own take on a concept of a made-for-TV interface. While it was used in the first-ever Apple TV and was itself a good reason to include a fancy iPod-like magnetic remote with the iMac G5, it lasted way less than Windows Media Center. In fact, as Front Row was dropped four years before the Microsoft solution, even SideShow managed to outlive it by two years.

It does feel strange to see the company the iPod and iTunes playing catch up with Microsoft on a media field. But it the end, it’s Apple and streaming services who ended up winning, not Microsoft’s pocket versions of the Media Center.

The core idea of a “media computer” as some sort of home entertainment server was pushed out of the mainstream by Netflix, Spotify, and other on-demand services. Once smartphones and mobile web became popular, the whole concept of a home PC as defined by the 1990s landscape folded in on itself. Finally, TVs and home appliances became more self-sufficient, and caught up with PCs in features which mattered. And while the overreliance on streaming comes with its own disadvantages, the public has made its choice.

This article from March 2019 was originally published on Yuri Litvinenko’s personal blog. A new video has been added, and the article was edited and restructured to better fit the 30pin style.

We all love a story of a major company releasing a dud, and the story of Microsoft and their Kin phones, discontinued in 2010 less than two months after the launch, certainly qualifies. It took a multitude of factors to make those phones fail: the feature set was not great, the competition was strong, and Verizon's plans were too steep for the intended audience.

That’s not to say the idea wasn’t wrong. Microsoft’s first attempt to be recognized as a phone manufacturer, made by creators of Sidekick phones, was built with the greater focus on social sharing—conceptually, at least. Its tile-filled start screen was more focused on social statuses rather than apps, and the Kin Spot widget, letting a user drop any content to share, was never found is any other Microsoft software.

The inside story of Kin’s downfall was documented within days of its formal discontinuation. Engadget sources put blame to a conflict between two Microsoft execs, while Business Insider revealed the demoralizing effect the failure of Kin caused on the company’s employees. But before that, Microsoft was interested in evolving their phones over time, and its software was meant to be more advanced than the one released. Last year, I talked to some people involved in creating the Kin lineup—and they shed more light on the vision behind Microsoft’s ill-fated project.

3D Interface and a Music Store Found on Public Videos

Microsoft’s marketing material, ads, and even phones’ own back covers proclaimed either Kin as a “Windows phone.” But it wasn’t powered by Windows Phone 7—that system was still in the making. Neither it was “Windows® phone” (sic), a moniker added to the last-ever revisions of the aging Windows Mobile 6 platform. Kin phones were running a system technically similar to either of platforms, but with a unique, custom-made user interface.

The UI, while ambitious and cool-looking, did not perform well. ComputerWorld’s Ginni Mies felt that navigating through Kin menus was “a bit sluggish”, while a video from PocketNow shows how choppy the animation was—even on a lock screen. The pre-release version responded poorly to swipes and touches as well, according to Microsoft’s study videos published by Wired in 2012. (The video, now deleted from Wired’s web site, was uploaded to Internet Archive.)

Most of the interactions and animations we designed ended up not being possible

With that in mind, it's hard to imagine Kin phones using a motion-rich 3D interface—but that’s what designers working on Kin envisioned, said Erik Hunter. Working for a design studio Method during the Kin development, he was collaborating its lead designers in the field of motion design.

“There were lots of interactions which didn't end being developed, like the motion on the lock screen, the Spot icon animation, and the scrolling tiles. Most of the interactions and animations we designed ended up not being possible,” Hunter said. The preliminary version of Kin’s user interface, looking notably different from the released one, is showcased on Hunter’s web site.

In the end, Kin phones did not have the same processing power and feature set which designers were using as a reference. "We were never given a reason for why the Spot wasn't as powerful as we originally designed it," Hunter added. He pointed to the internal conflict between Microsoft's managers and multiple mobile teams, reported by Engadget at the time, but did not confirm it from his own perspective.

The vision had to be downscaled. Chris Furniss, who was implementing the interface in XAML and Silverlight (the same markup and framework used in Windows Phone), was one of people tasked with that. He noted that a chunk of the UI had to be redesigned, as the change in animations led to a chain reaction within other elements.

“The UI, in general, had a cool parallax effect as you swiped back and forth, or scrolled up and down through your social feed. That had to be cut, which affected the way text had to be layered over photos and so forth,” Furniss explained.

The eye candy wasn't the only thing that had to be cut: the Spot, one of Kin’s unique features, ended up being less functional than originally outlined. In its original form, Kin Spot was “a context-sensitive bucket which you could drop anything into and Kin would figure what you were intending.” Furniss recalled.

Portions of that vision are seen in Hunter’s showcase. The Kin Spot, being just a drag-and-drop target for SMS, MMS, and email attachments, is more advanced in his video. Notably, it features a Spaces section, not seen on the retail version of Kin phones. Within Spaces, according to the video, a user could collect objects across the system to folders, creating what looked like custom feeds.

Another video, uploaded in 2011 by Microsoft’s Joseph McLaughlin (under his Aesthetikal creative label), shows a Kin interface which looks almost identical to the version shipped, but with richer animations, parallax scrolling within tiles, and a three-dimensional perspective effect while scrolling in both directions.

Just like the other video, it shows features considered by developers but not included in the final product. A Windows Live Messenger chat screen is especially striking, considering that Kin shipped with no IM support at all. The other feature not seen on actual phones is the media marketplace, integrated within the main UI and getting the featured placement in a bottom left corner. Instead of that, Kin phones had a carbon copy of Microsoft’s Zune HD media player software as a separate app. (30pin has reached to McLaughlin for comment on Twitter and will update the article upon getting a reply.)

The decision to effectively port Zune software to Kin phones, complete with a clashing Metro design language, mimics the choice of hardware: Kin phones used the same Tegra APX 2600 chip by Nvidia marked the first time Tegra, Nvidia’s system-on-a-chip, was used in a mobile phone. While that technically makes that first-ever phones powered by Tegra, the exact chip was almost two years old by then, with the successive Tegra 2 lineup already announced by Kin’s release.

“Kin Three,” the One for Grown-Ups

Microsoft marketed Kin phones to members of what it called an “upload generation,” with teens and tweens in ads and a brief, quickly-pulled sexting scene in one of them. The definite age of projected demographic varied by model, Furniss noted. Kin Two, the top-of-the-line model which looked like a contemporary QWERTY side slider, was aimed at people between 25 and 35, stretching the definition of youth a bit. The cheaper, oddly-shaped Kin One was the one meant for pre-teens and teens—and, according to Furniss, it was specifically designed to appeal to the female audience.

But looking forward, Microsoft and its partners were already planning to reach another demographic. One developer, who decided to remain anonymous due to NDAs, shared that the idea was to keep people captured by first-generation models even as they grow up and with their tastes changing.

“If Microsoft Kin is a mobile device designed specifically for a younger, millennial audience, our job was to think about how the device needs to change as this audience becomes older,” they describe their role in the project.  The source states their team researched the potential new audience and outlined new features for the future Kin.

Several Kin developers shared the project was meant to be something more than an one-off venture by Microsoft. “We were all in creative brainstorm mode for the next big thing, and spirits were high,” recalls Furniss the mood across his team. Hunter added that he and his team at Method “given lots of room to explore” and “encouraged to push the boundaries of what can be done in a traditional mobile UI.”

Alas, there was no other phones in the series. Microsoft embarrassingly discontinued Kin phones after six weeks of low sales. Later in 2010, phones were stripped of social and cloud aspects, re-released with an “m” suffix, stand effectively put against cheaper basic phones. Those who did get original models were left only few more month to use the cloud, as Verizon closed the cornerstone Kin Studio service next February.

Kin phones were released in the most unlikely of times for Microsoft's mobile divisions. There were lofty expectations for all-new Windows Phone; the Kin software, having no app support, fared worse compared to it and other smartphone systems.

“When Kin was being designed, it was before this huge paradigm shift to consuming everything via platform-specific apps,” noted Furniss. Kin phones bet on integrated experiences, not apps—but to win that bet, Kin had to either be as refined as initially planned or cost as much as it deserved to. Unfortunately for everyone involved in its creation, it did neither.

This article from February 2019 was originally published on The Conversation. It’s released under a Creative Commons Attribution-No Derivatives 4.0 license which allows republishing. It has been slightly updated to reflect changes in time.

Five years ago, I wrote a history of 160 years of faxing, saying my book covered “the rise and fall of the fax machine.” The end I predicted has not yet come: Millions of people, businesses and community groups send millions of faxed pages every day, from standalone fax machines, multifunction printers and computer-based fax services. It turns out that in many cases, faxing is more secure, easier to use and better suited to existing work habits than computer-based messaging.

Businesses Often Use Faxes

Faxing remains alive and well, especially in Japan and Germany—and in major sectors of the U.S. economy, such as health care and financial services. Countless emails flash back and forth, but millions of faxes travel the world daily too.

A worldwide survey in 2017 found that of 200 large firms, defined as companies with more than 500 employees, 82 percent had seen workers send the same number of, or even more, faxes that year than in 2016. A March 2017 unscientific survey of 1,513 members of an online forum for information technology professionals found that 89 percent of them still sent faxes.

The persistence of faxing—and the people who send faxes—is in part because the fax industry has adapted to accommodate new technologies. Fax machines still dominate, but both surveys suggested users were shifting to computer-based services, such as fax servers that let users send and receive faxes as electronic documents. Cloud-based fax services, which treat faxes as images or PDF files attached to emails, are also becoming more popular. These new systems can transmit faxes over telephone lines or the internet, depending on the recipient, handling paper and electronic documents equally easily.

Legal Acceptance

Fax’s longevity also benefits greatly from reluctance—both legal and social—to accept email as secure and an emailed electronic signature as valid. Faxed signatures became legally accepted in the late 1980s and early 1990s in a series of legal and administrative decisions by state and federal agencies. The Electronic Signatures Act in 2000 also gave digital signatures legal power but institutional and individual acceptance followed only slowly—if at all.

Even parts of the federal government preferred faxes over email for many years thereafter. Not until 2010 did the Drug Enforcement Agency allow electronic signatures for Schedule II drugs like Ritalin and opiates, which comprised about 10 percent of all prescriptions. That meant a pharmacist could accept a faxed prescription but not one scanned and sent by email.

The most recent FBI Criminal Justice Information Services policy allows faxing from physical fax machines without encrypting the message, but demands encryption for all email and internet communications, including cloud-based faxing. It’s much harder to intercept faxes than unencrypted email messages.

Faxing and Medicine

Another reason faxing hangs on is because competing technologies are weak. The health care industry generates huge amounts of data for each patient. That should make it fertile ground for a fully digital record-keeping system, “where data can flow easily between patient, provider, caregivers, researchers, innovators and payers,” as Seema Verma, the head of the Centers for Medicare and Medicaid Services, put it in a speech earlier this year.

Federal privacy laws and deliberately incompatible standards, however, stand in the way. Immediately after the passage of the 1996 Health Insurance Portability and Accountability Act, fax vendors retooled their transmission, reception and storage systems and procedures to protect patients’ personal records. Specifically, HIPAA-compliant fax systems ensure the correct number is dialed and limit who can see received faxes. Digital patient-information systems have struggled to meet the same standards of administrative, technical, and physical security.

The Obama administration spent more than $25 billion encouraging doctors and hospitals to adopt electronic medical records systems. Crucially, rather than forcing competing systems to be compatible in order to receive federal support, the administration believed the market would decide on a standard to communicate.

What actually happened was that competing companies deliberately created incompatible systems. Doctors’ offices and hospitals that use different records databases can’t communicate with each other digitally – but they can via fax. For many medical professionals, particularly independent physicians, faxing is far easier than dealing with expensive, hard-to-use software that doesn’t actually do what it was supposed to: securely share patient information.

Comfortable Inertia

One more personal factor preserving faxing is users’ reluctance to change. Small businesses who find that faxing meets all their needs have little reason to spend the money and effort to try a new technology for document exchange. Every company that prefers faxes inherently encourages all its customers and suppliers to keep faxing too, to avoid messing up existing ordering processes.

It’s important to remember, too, that fax machines and multifunctional printers with a fax capability provide an inexpensive backup capability in case of technical problems with an internet connection, or even a cyberattack, like the Russian attack on Estonia in 2007.

Absent a compelling reason or some management or government mandate, people often don’t change technologies. This is true beyond faxing: I drive a 2005 Camry. There are plenty of cars that are better in some way – safer, more fuel-efficient, more comfortable – but so long as the Camry passes state inspections and performs adequately, I can avoid the challenges and costs of buying a new car and learning how to use its new features.

International Popularity

Faxing is still popular overseas, too. In Britain, the 2000 Electronic Communications Act encouraged but did not explicitly authorize electronic signatures. In 2018, urged partly by the European Union’s promotion of electronic identification, the British Law Commission concluded that electronic signatures were indeed legal but needed significant promotion to increase their acceptance and use.

Not surprisingly, a 2018 survey found that Britain’s National Health Service operated more than 8,000 fax machines. In response, the U.K.’s Secretary of State for Health and Social Care Matt Hancock labeled faxing a symbol of National Health Service technological backwardness and pledged to introduce new technologies more quickly. In December, the National Health Service decided to stop buying fax machines in 2019 and end their use by the end of 2020. That’s the same goal the Centers for Medicare and Medicaid Services’ Verma has for American doctors to stop faxing.

Nevertheless, faxing continues because it remains better – cheaper, more convenient, more secure, more comfortable – than the alternatives for many people, businesses and organizations. Faxing will remain important until transmitting digital data becomes easier and more accepted. That could be a long way off, though. U.S. federal initiatives are trying to make medical records systems more compatible, but no one has yet been hired to take a key leadership position at CMS.

Eventually the older generation of people more comfortable with faxing than emailing will fade away. Until then, however, fax machines will whirl away.