Category: Interesting

  • How the UK Games Industry Moved Online: From Teletext Cheats to Early Download Stores

    How the UK Games Industry Moved Online: From Teletext Cheats to Early Download Stores

    There is a particular kind of nostalgia attached to British gaming that has nothing to do with the games themselves. I am thinking of the Saturday morning ritual: kneeling in front of the television at an ungodly hour, notebook in hand, waiting for Teletext page 150 to cycle round to the cheats section. If you blinked at the wrong moment, you missed it, and the whole carousel started again. That was the internet before the internet, and for several years it was entirely sufficient.

    The UK games industry online history is, in many ways, the story of an audience that was always a few steps ahead of the infrastructure available to it. British players wanted more, more cheats, more previews, more contact with developers, long before broadband made any of that genuinely practical. What filled the gap was a patchwork of surprisingly ingenious solutions: premium-rate phone lines, fax-back services, magazine cover discs, and yes, Ceefax and its commercial rival Oracle. The path from those stop-gap measures to Steam’s launch in 2003 is a shorter journey than it looks, but it passes through some genuinely fascinating terrain.

    Old CRT television showing a Teletext-style cheats page, part of UK games industry online history
    Photo by Anete Lusina on Pexels

    Teletext cheats pages and the premium tips line era

    Teletext’s games coverage was not especially deep, but it was instantaneous by the standards of the early 1990s. Channel 4’s Teletext service (the commercial successor to Oracle) ran dedicated gaming sections that listed cheat codes, level passwords, and occasionally brief reviews. The BBC’s own Ceefax service carried similar material. Neither was interactive, and neither let you ask a follow-up question. That gap was filled, lucratively, by premium-rate telephone tips lines.

    Games companies and third-party operators set up dedicated 0891 numbers throughout the late 1980s and into the mid-1990s. Calls were charged at up to 50p per minute, and a single session talking to an automated system, or, on better lines, an actual human being with a ring-binder full of solution guides, could run to several pounds. The BT quarterly bill arriving with a mysterious £30 addition caused genuine family arguments across Britain. I have spoken to people who still wince at the memory. According to period coverage in Edge magazine, some lines were generating tens of thousands of calls per week at the height of the Super Nintendo era.

    What the tips lines showed, beyond their obvious profitability, was that British players had an appetite for information they were prepared to pay for. That appetite did not disappear when the internet arrived. It simply migrated.

    The magazine and cover disc bridge

    Before dedicated gaming websites existed, the UK had something arguably better: a games press that was the finest in the world. CRASH, Zzap!64, Mean Machines, CVG, GamesMaster and eventually Edge formed a publishing ecosystem that had no real equivalent abroad. These titles did not just review games; they explained them, satirised them, and built communities around them through letters pages that functioned as a primitive social network.

    The cover disc was the physical equivalent of a download. From the early 1990s onwards, magazines bundled floppy discs and later CD-ROMs containing demos, shareware, and occasionally full games. PC Zone and PC Gamer UK made cover discs a central part of their proposition. The CD-ROM format, in particular, carried enormous amounts of data by the standards of dial-up connections, meaning that players without fast internet could still access substantial trial content. The disc was, effectively, the offline download store.

    Stack of 1990s UK gaming magazines with cover discs, illustrating the UK games industry online history
    Photo by Roberto Lee Cortes on Pexels

    Early UK gaming websites and the dial-up era

    The UK games industry online history begins in earnest somewhere around 1994 to 1996, when individual studios and enthusiast communities started staking out territory on the web. The process was chaotic. Some of the most important early sites were run by fans rather than professionals, hosted on university servers via the same dial-up connections that everyone else was crawling along on.

    Rare, the Twycross-based developer behind Donkey Kong Country and GoldenEye 007, had an early web presence that was sparse by any modern measure but remarkable for its time. Core Design in Derby, the studio that gave the world Lara Croft in 1996, maintained a site that included developer diaries and fan interaction. These were not sophisticated operations. They were often a few static HTML pages maintained by someone who had taught themselves to code in an evening. But they worked, and they demonstrated that the audience was genuinely interested in the people behind the games, not just the games themselves.

    Eurogamer launched in 1999, initially as a small UK-focused review site. GameFAQs, though American in origin, built a British user base that had been trained by years of Teletext cheats pages to seek out exactly that kind of reference material. Meanwhile, the dot-com boom brought speculative investment into online gaming media, with ventures like Gameplay.com attempting to build integrated retail and content operations before the crash made that kind of ambition briefly unfashionable.

    Multiplayer servers and the emergence of online communities

    Quake, released in 1996, changed British gaming culture in a way that is difficult to overstate. id Software’s game shipped with built-in TCP/IP support, which meant that anyone with a modem and enough patience could connect to a server and play against other human beings. The experience, on a 28.8k modem, was not exactly smooth. Lag was a constant companion. But the basic idea, that you could play a game with someone in Edinburgh while sitting in Southampton, was genuinely new to most players.

    British gaming communities assembled around game servers maintained by enthusiasts, often on university networks where bandwidth was relatively generous. Clan culture developed rapidly. The early UK competitive scene for games like Quake and then Quake II was organised largely through mailing lists, IRC channels, and rudimentary league websites. The Clanbase network, which organised online competitions across Europe, drew substantial British participation throughout the late 1990s and early 2000s.

    Some of the digital communities that formed around multiplayer gaming at this period bear comparison to other kinds of online political and cultural organising. Researchers at activism.digital who study how internet communities form around shared causes will recognise the structural similarities: a dispersed group with a common interest, a shared information infrastructure, and the gradual development of norms and leadership. Games clans were not political, but they were organised, and they were distinctly British in their habits and humour.

    From download demos to early digital retail

    The transition to digital distribution happened in stages. The demo download was the first. By the late 1990s, studios were releasing timed or limited versions of their games as free downloads, a practice that had existed on the PC platform since the shareware model of the early 1990s but which now reached a wider audience through web hosting rather than floppy disc distribution.

    Gameplay.com and Jungle.com were early attempts at UK online games retail, selling physical boxes rather than downloads, but getting players used to the idea of ordering games without visiting a shop. The real shift came with Valve’s Steam platform, launched in September 2003. Steam’s initial rollout was not universally celebrated; the requirement to authenticate Half-Life 2 online before playing it caused considerable complaint. But within a few years, the model had become standard, and the British high street games shop was on notice.

    What the UK brought to online gaming

    Britain’s contribution to this story is not simply that of a passive consumer market. The UK games press trained a generation of players to think critically about games at a time when most countries treated them as pure product. Studios like Bullfrog Productions in Guildford, Lionhead in Guildhead, and Rockstar North in Edinburgh built international reputations while remaining distinctly British in sensibility. The dry wit of Theme Hospital, the moral ambiguity of early Grand Theft Auto, the philosophical ambition of Black & White: these were not accidental qualities.

    The UK games industry online history is also the history of how a particular kind of community knowledge, the sort of thing that once lived in a Teletext page or a tips-line recording, found its permanent home on the web. It did not happen overnight, and the path was less linear than retrospective accounts often suggest. But the destination, a fully networked games culture with British studios, British press, and British communities at its centre, was always where that Saturday morning ritual with the notebook was pointing.

    The cheats pages are gone. The tips lines are long since disconnected. What replaced them turned out to be considerably larger.

  • The Quiet Death of AltaVista UK: How British Users Lost Their First Search Engine of Choice

    The Quiet Death of AltaVista UK: How British Users Lost Their First Search Engine of Choice

    There was a moment, somewhere around 1997 or 1998, when typing a question into AltaVista felt genuinely miraculous. The site had indexed what seemed like the entire known web, it returned results in seconds over a dial-up connection, and it understood proper search syntax that most people had never seen before. For millions of British users, it was the first real search engine they ever used. I’d argue it shaped how an entire generation thought about finding things online, and its disappearance was so gradual that most people simply didn’t notice it was gone until years later.

    The AltaVista UK search engine history is not quite the same story as its American trajectory. British users came to the web under very different circumstances, through ISPs bundling their own portals, through Ceefax-adjacent curiosity, and through a telecommunications landscape where every minute online cost money. The search engine you used was often the one your ISP had pre-selected for you, and AltaVista managed, for a few extraordinary years, to be genuinely dominant even within that constrained environment.

    Close-up of an early web browser search box, illustrating AltaVista UK search engine history
    Photo by Matheus Bertelli on Pexels

    How AltaVista actually reached British homes

    AltaVista launched in December 1994, a product of Digital Equipment Corporation’s research lab in Palo Alto. By 1996 it was the most visited site on the web globally, handling around 80 million queries per day at its peak. In Britain, the path to dominance was more winding. Most households were still connecting through services like Demon Internet, CompuServe or the early wave of free ISPs that Freeserve helped make standard. These services often pushed users towards their own start pages, with curated directories of “best of the web” links that felt more like a magazine contents page than a search box.

    AltaVista survived this portal era by being genuinely better than the alternatives. Yahoo’s directory was human-curated and slow to update. Lycos returned too much noise. Excite and Infoseek had their advocates, but AltaVista’s index was simply larger and its results more precise. British users discovered that if you wanted to find something specific, such as a football result, the lyrics to a B-side, or the address of a local business, AltaVista was the place to go. Its advanced search operators, Boolean queries with AND, OR and NOT, gave technically minded users a tool that felt almost professional. I’ve spoken to people who remember printing out AltaVista search tips and keeping them next to the family computer.

    The British web directories that shaped what people searched for

    UK internet culture in the late 1990s had a particular character that is easy to forget now. The web felt small enough to catalogue. Sites like Yell.com were just finding their feet. UK Plus, a British web directory that launched in 1995, was widely used as a complement to search engines rather than a replacement. The BBC maintained its own curated web guide. BT Internet and Virgin Net both offered portal pages with pre-selected links intended to keep users inside a walled garden of approved content.

    AltaVista’s great advantage was that it pointed outward. It acknowledged the chaos of the web and tried to map it, rather than taming it into a directory. British users who outgrew their ISP’s start page tended to drift towards AltaVista as their default, particularly those in universities who had faster connections and more time to explore. The dial-up era’s metered cost structure actually reinforced this habit: you wanted results fast, from a source you trusted, and you didn’t want to spend your online minutes hunting around.

    When the wheels started coming off

    The decline of AltaVista is usually dated to its acquisition by CMGI in 1999, followed by a disastrous attempt to transform it into a full-service internet portal with shopping, email, news and entertainment. This was precisely the wrong strategy at precisely the wrong time. AltaVista’s strength was search. Diluting that with a cluttered homepage full of competing links made it slower and harder to use, at the exact moment a cleaner, faster alternative was emerging.

    Google’s UK presence grew quietly at first. The company launched Google.co.uk in 2000, and British users began switching through a kind of word-of-mouth that is hard to recreate today. I’ve heard it described as a moment of sudden clarity: you tried Google once, the results page was clean and the results were better, and you simply never went back to AltaVista. The PageRank algorithm, which ranked pages by the quality and quantity of links pointing to them, happened to suit the British web particularly well. UK websites had been building genuine editorial link structures through years of web directories and recommendation culture, and Google rewarded exactly that.

    Meanwhile AltaVista lurched through a series of ownership changes. Overture Services acquired it in 2003. Yahoo bought Overture later the same year, folding AltaVista into an empire that was itself struggling to compete with Google. The dot-com crash of 2000 to 2001 had stripped away much of the speculative investment that had kept second-tier search engines afloat, and AltaVista never recovered the development momentum it would have needed to close the gap with Google’s rapidly improving index.

    ISP tie-ins and why British users were slow to leave

    One underappreciated reason AltaVista held on longer in Britain than elsewhere was the ISP portal system. Freeserve, BT Internet and other providers had deals or informal alignments that kept AltaVista on recommended lists and browser toolbars well into the early 2000s. If your ISP’s start page had an AltaVista search box embedded in it, and if you were a casual user who simply wanted to find something quickly, you might not have noticed that better options existed. This is not so different from how search defaults work today: the engine you see first tends to be the engine you use, which is why the question of default search agreements still matters enormously to regulators and to the industry.

    For anyone thinking about how search engines historically gained dominance through these kinds of tie-ins, it is worth understanding what that means for visibility today. A site that ranked well on AltaVista in 1999 was not necessarily visible on Google in 2003 without conscious effort. The metrics were different, the crawlers behaved differently, and the whole framework of what made a page findable had shifted. UK businesses that had optimised for AltaVista’s keyword-heavy approach often found themselves invisible on Google until they understood PageRank. The principle of checking how your site performs against current search criteria remains relevant: Search Engine Tuning, a UK-based tool that lets website owners run a free SEO check across their domains, represents the modern version of what webmasters in 2003 were frantically trying to work out manually. At searchenginetuning.co.uk, you can check your SEO standing against the signals that Google actually uses today, rather than the keyword stuffing that AltaVista once rewarded.

    What was lost when AltaVista went dark

    AltaVista was officially shut down on 8 July 2013, by which point it had become a footnote even within Yahoo’s own portfolio. The BBC reported the closure with a kind of amused nostalgia, noting that the engine had once indexed 100 million web pages at a time when that felt like everything. For most British internet users, the closure confirmed something that had effectively happened a decade earlier.

    What the AltaVista UK search engine history really records is the transition from a human-scaled web to something too large for any single index to claim comprehensively. In the late 1990s, the web felt like a library where you might actually browse the whole collection. AltaVista was the card catalogue. When that illusion broke, and when Google’s infrastructure made it clear that no human-readable index could keep pace, AltaVista’s founding premise became obsolete almost overnight.

    The engine also leaves behind a specific legacy in search culture. Its advanced operators, the plus and minus signs, the quotation marks for exact phrases, all migrated into Google’s own interface where many still work today. British researchers who learnt to search on AltaVista carried those habits forward, and the muscle memory of precise Boolean querying is something librarians and academics still teach. According to the British Library’s own research guidance, structured query techniques remain a core skill for academic search, even if the engine underneath has changed beyond recognition.

    The SEO habits that AltaVista created were harder to shift. Many UK websites built in the late 1990s were optimised for keyword density rather than link authority. When Google took over, those sites discovered their past efforts counted for little or nothing. The lesson, then as now, is that search engine visibility depends entirely on understanding which signals the dominant engine actually weighs. Checking your domains against current criteria is precisely what specialists like Search Engine Tuning exist to help with: their free SEO check scans a site against the ranking signals that matter to Google in 2026, not the keyword-counting approaches that AltaVista once rewarded.

    The quiet death of AltaVista is, in the end, a story about how quickly the internet forgets its own infrastructure. The engine that taught Britain to search is now accessible only through the Internet Archive, its results pages preserved in amber, still returning links to sites that have themselves long since disappeared. There is something genuinely moving about that. A search engine whose entire purpose was to find things, now findable only by those who already know where to look.

  • What the Dial-Up Handshake Actually Meant: A Technical and Cultural History of UK Home Internet Before Broadband

    What the Dial-Up Handshake Actually Meant: A Technical and Cultural History of UK Home Internet Before Broadband

    There was a sound that defined a generation of British households, and it came from the spare bedroom, or the corner of the living room, or wherever the family computer happened to live. A screech, a hiss, a sequence of tones that resembled nothing so much as two fax machines having an argument. Then silence. Then, if you were lucky, the words Connected at 44,000 bps appearing in a small grey box on screen. That was the dial-up handshake. And for most British families between roughly 1995 and 2004, it was the sound of the internet arriving.

    I’ve spent a good while thinking about what dial-up actually was, technically and socially, because I think we’ve misremembered it. We treat it as a joke now. A punchline about loading times and patience. But it was genuinely complex technology grafted onto infrastructure never designed to carry it, and the social structures it imposed on households were stranger and more specific than most histories acknowledge.

    Late 1990s desktop PC with 56k modem representing dial-up internet UK history before broadband
    Photo by MART PRODUCTION on Pexels

    How a telephone line carried the internet

    The public switched telephone network in Britain was built for voice. BT’s copper wire infrastructure, much of it laid decades before the web existed, was designed to carry analogue audio signals in a frequency range of roughly 300 Hz to 3,400 Hz. That is the range human speech occupies. The modem’s job, and the word itself is a contraction of modulator-demodulator, was to squeeze digital data into that narrow audio channel.

    Early modems in the late 1980s managed this at 2,400 bits per second. By the mid-1990s, the V.34 standard pushed that to 33,600 bps. The 56k modem, formally standardised as V.90 by the ITU in 1998, was the pinnacle. It could theoretically receive data at 56,000 bits per second, though in practice British BT lines rarely delivered more than 44,000 to 48,000 bps. The upload speed was capped lower still, at 33,600 bps, because the asymmetric design assumed you’d be downloading more than you sent.

    The handshake itself, that extraordinary cacophony, was a genuine negotiation. The calling modem and the answering modem at your ISP’s point of presence were exchanging training sequences, testing the line quality, agreeing on a protocol and settling on the fastest speed the line could reliably sustain. Different tones corresponded to different phases of this process. The initial high-pitched squeal was carrier detection. The lower rumbling tones were the training sequence. The brief silence before the connected message was the moment both modems agreed terms. It took, on a good day, about eight seconds. On a bad line, closer to thirty.

    The BT phone bill problem and off-peak browsing culture

    Here is the thing that younger readers find genuinely difficult to believe: every minute you spent online cost money from your phone bill. BT charged local call rates for dial-up connections, and in the mid-1990s that meant roughly 4p per minute at peak times, dropping to about 1.5p per minute during evenings and weekends. A single evening of browsing could add several pounds to the quarterly bill. An enthusiastic teenager could, and frequently did, add tens of pounds without realising it.

    The result was a very British form of internet rationing. Households developed unspoken rules, sometimes spoken ones, about when and how long anyone could use the internet. Off-peak hours, after 6pm on weekdays and all weekend, became the acceptable window. Many families set timers. Some parents instituted outright bans during peak hours. I’ve heard accounts from people who would queue their web pages, literally writing down URLs on paper during the day, then connect in the evening to load them all in sequence before disconnecting as quickly as possible.

    The free ISP revolution, which you can read about in detail in the history of Freeserve and how Dixons triggered Britain’s free ISP era, eliminated monthly subscription fees from around 1998. But it did nothing about the phone charges. Those persisted until BT launched its own unmetered service in 2000, after considerable political pressure and competition from alternatives. Even then, unmetered dial-up was a separate paid product. The myth that the internet was ever truly free during the dial-up era is exactly that: a myth.

    Tying up the phone line and family negotiations

    The telephone occupied a peculiar status in British households of the 1990s. It was infrastructure. People expected it to work when they picked it up. The arrival of a household modem introduced a conflict that had no precedent: the telephone line could only do one thing at a time, and a connected modem made the line completely unavailable for voice calls.

    The social consequences were significant. Anyone trying to ring the house got an engaged tone. Anyone in the house wanting to make a call had to either wait or ask the internet user to disconnect. Family arguments about this were so common they became a cultural reference point. BT eventually offered a product called Home Highway, a form of ISDN that gave households two simultaneous channels on one line, allowing voice and data at the same time. It was expensive, and relatively few families took it up.

    Call waiting services created their own complications. Some modems could detect a call waiting signal and drop the connection automatically. Others would corrupt the data stream when the tone arrived. The technical solution was to dial *43# before connecting, which disabled call waiting for the duration of the session. Many households discovered this trick by accident, after one too many disconnections mid-download.

    There was also the question of what happened when someone picked up an extension. Lifting a handset in another room whilst the modem was connected introduced noise onto the line. At low speeds this was merely annoying. At 56k, it could drop the connection entirely. The family rule “don’t pick up the phone when someone’s on the internet” became as standard as “don’t slam the front door.”

    What you could actually do at 56k

    The speed debate gets misremembered as straightforwardly slow. Dial-up was slow compared to what followed. But the early web was also built for dial-up. Text loaded quickly. Simple images loaded in a few seconds. A basic HTML page from 1998 might total 30 to 50 kilobytes; at 40,000 bps, that’s around ten seconds. Manageable.

    The problem was what the web became. Image-heavy pages, increasingly common from 2000 onwards, could run to several hundred kilobytes. A JPEG photograph at reasonable resolution was 80 to 150 KB on its own. Audio files were largely impractical to stream; you downloaded them and played them locally, which was partly why the music piracy era, covered in the history of UK online music piracy and Napster’s impact on British listeners, involved so much waiting. A three-minute MP3 at 128 kbps was roughly 2.8 MB. At 40,000 bps, that’s around nine minutes of downloading, assuming a stable connection throughout.

    Video was essentially impossible. Flash animations, the kind that took over British school computer rooms in the late 1990s, could take ten minutes to load for a thirty-second clip. Most people waited. The progress bar became an object of meditation.

    The end of dial-up and what replaced it

    BT’s ADSL rollout, which began seriously around 2000 and accelerated through 2002 and 2003, changed everything about the technical relationship between British households and the internet. ADSL, Asymmetric Digital Subscriber Line, used the same copper telephone wire as dial-up but carried data on frequencies above the voice range, leaving both channels available simultaneously. You could be online and receive a phone call at the same time. The connection was permanent. There was no handshake, no engaged tone, no off-peak calculation.

    The early ADSL speeds, typically 512 kbps downstream, were roughly ten times faster than a good dial-up connection. By 2005, 2 Mbps packages were common. The social rituals of dial-up, the timers, the queued URLs, the arguments about the phone line, dissolved within a year or two of each household switching. The habits were so deeply ingrained that some families kept them anyway for a while, through sheer conditioning.

    According to BBC reporting on the history of UK broadband adoption, dial-up connections still accounted for a significant share of UK home internet access as late as 2005, with the transition to broadband taking longer in rural areas where ADSL infrastructure arrived years behind urban centres.

    The dial-up era lasted roughly a decade in British homes. It was ungainly, expensive by modern standards, socially disruptive in ways nobody had anticipated, and technically remarkable given the constraints it worked within. The handshake that opened every session was a genuine feat of signal engineering. The rules that grew up around it were a genuine feat of household diplomacy. Both deserve to be remembered with something more than embarrassment.

    Frequently Asked Questions

    How fast was 56k dial-up internet in the UK really?

    In theory, V.90 modems could receive data at up to 56,000 bits per second. In practice, UK BT telephone lines, which were analogue copper infrastructure, typically supported connections of 44,000 to 48,000 bps. Upload speeds were capped lower, around 33,600 bps, by the asymmetric design of the V.90 standard.

    Why did dial-up internet make your phone line busy?

    Dial-up worked by using your standard BT telephone line to make a call to your ISP’s server. Because the line was occupied making that call, no other calls could come in or go out whilst you were connected. It was technically identical to making a voice call, except the audio was modem data rather than speech.

    How much did dial-up internet cost in the UK in the 1990s?

    The cost came primarily from BT’s local call charges rather than ISP subscription fees, particularly after free ISPs arrived in 1998. During peak hours, BT charged roughly 4p per minute for local calls; off-peak rates dropped to around 1.5p per minute. A heavy session of two hours in the evening could therefore cost around £1.80 in call charges alone.

  • Inside the UK Dot-Com Boom: The City Investors, Shoreditch Start-Ups and Media Frenzy of 1999 to 2001

    Inside the UK Dot-Com Boom: The City Investors, Shoreditch Start-Ups and Media Frenzy of 1999 to 2001

    There is a particular kind of collective madness that descends on financial centres every few decades. Tulip bulbs in seventeenth-century Amsterdam. Railway mania in Victorian Britain. And then, at the very end of the twentieth century, the internet. I’ve spent a long time reading through the press archives from 1999 and early 2000, and what strikes me most is not the greed, it’s the absolute sincerity. People genuinely believed the old rules of business had been repealed. Profit was a quaint Victorian notion. The UK dot-com boom 1999 2001 history is, in many ways, a story about a nation convincing itself that this time really was different.

    City of London financial district, the source of venture capital during the UK dot-com boom 1999 2001 history
    Photo by Chengxin Zhao on Pexels

    The Square Mile pours money eastward

    By 1998, the City of London was watching San Francisco with barely concealed envy. American venture capital firms were minting millionaires at a rate that made traditional banking look like a retirement hobby. British institutional investors, not wanting to be left behind, began routing serious capital into a cluster of start-ups that had gathered, somewhat improbably, around Old Street roundabout in east London. Estate agents were already calling it Silicon Roundabout, a nickname that would stick for another two decades.

    The numbers got absurd very quickly. According to research compiled at the time by the British Venture Capital Association, VC investment in UK technology companies more than trebled between 1997 and 1999. The City wasn’t just curious; it was throwing money at anything with a .com suffix and a presentation deck. I’ve seen some of those original pitch documents, they are extraordinary artefacts. Revenue projections drawn as hockey sticks. Business models that essentially said “we will acquire users, and monetisation will follow.” The investors nodded and wrote cheques.

    The companies that defined the bubble

    A few names from this period deserve particular attention, because they capture just how far the mania extended into British commercial life.

    Boo.com is the one everyone remembers. Founded in 1998 by Swedes Ernst Malmsten and Kajsa Leander, but headquartered in London and funded largely by British and European money, Boo burned through roughly £80 million in under two years trying to build a global online fashion retailer before the technology infrastructure existed to support it. The site required a broadband connection to run properly at a time when most British households were still on dial-up. It collapsed in May 2000, and its domain name sold for £1.2 million. The story of Boo.com became shorthand, in the financial press, for everything that had gone wrong.

    Then there was Clickmango.com, a health and beauty e-commerce site that raised £6 million and burned through it in months. Gameplay.com, which sold games online and actually had a functioning business model, found itself valued at figures that bore no relationship to its revenues simply because it existed on the internet. QXL Ricardo, a British rival to eBay, raised vast sums on the back of auction mania, you can read more about how that particular corner of the web played out in the history of UK online culture in the late 1990s, where the same feverish energy was reshaping everything from music distribution to retail.

    The media made it worse

    Journalism does not cover itself in glory during bubble periods. The broadsheets ran breathless profiles of twenty-six-year-old CEOs who had never run anything larger than a university society. The Evening Standard published a “dotcom rich list” tracking the paper fortunes of British internet entrepreneurs. Television followed. Channel 4 commissioned programmes about the new economy. The BBC ran features on families getting rich from their garages.

    I find the newspaper archives from 1999 genuinely difficult to read without wincing. The financial pages are full of phrases like “first-mover advantage” and “land grab” and “eyeballs”, that peculiar metric by which companies were valued not on earnings but on the number of people looking at their website. The logic was circular and everyone knew it, but the music was playing and nobody wanted to sit down.

    The Ananova story sits somewhere in this moment, too. The world’s first digital news presenter, built in Leeds by PA New Media, launched in April 2000 just as the NASDAQ was beginning its long collapse. It was exactly the kind of forward-looking digital project that had seemed self-evidently brilliant eighteen months earlier. By the time it launched, the mood had already shifted.

    What the crash actually looked like on the ground

    The NASDAQ peaked on 10 March 2000 at 5,048 points. By October 2002 it had fallen to 1,114. The decline was not immediate in British boardrooms, there was a lag of a few months, a period of collective denial, before the redundancy notices started appearing. I’ve spoken to people who were working in Shoreditch start-ups in the summer of 2000, and many of them describe a strange limbo: the money had stopped, the valuations had collapsed, but the offices were still open and the ping-pong tables were still there.

    Then, quite suddenly, they weren’t. Entire companies vanished over weekends. Staff arrived on Monday mornings to find the doors locked. The Guardian ran a jobs section specifically for displaced dot-com workers. Recruitment agencies reported a glut of “internet marketing managers” and “community managers”, job titles that had barely existed three years earlier and now described thousands of people looking for work.

    A detailed account of which specific British start-ups went under and how the City processed those losses is something I’ve written about separately. The collapse of Britain’s dot-com start-ups reads, in hindsight, like a very expensive lesson in what happens when capital chases narrative rather than fundamentals.

    What survived and why

    Not everything died. The companies that made it through the crash shared a quality that is obvious in retrospect but was surprisingly unfashionable during the boom: they had revenues. Lastminute.com floated in March 2000 at the absolute top of the market, saw its share price collapse by more than 90%, but kept trading because it was selling actual airline tickets and hotel rooms. It eventually sold to Sabre Holdings in 2005 for £577 million. Not the fantasy valuation of its flotation, but real money for a real business.

    The infrastructure companies fared better than the consumer-facing ones. The ISPs, the data centres, the telecoms firms laying fibre, these had assets and customers and recurring revenue. It’s part of why the story of early British internet history is, ultimately, one of infrastructure rather than innovation. The Freeserve era had already demonstrated that the real money in British internet history often lay not in the glamorous front-end start-up but in the plumbing underneath it.

    The UK dot-com boom 1999 2001 history is worth studying precisely because it rhymes so clearly with later cycles. The same patterns reappeared during the social media investment bubble of the late 2000s, and again during the crypto mania of the early 2020s. Capital floods in, valuations detach from reality, a trigger event causes a cascade, and then the post-mortem begins. Britain’s version of the dot-com crash was, in some ways, more contained than America’s simply because the British VC ecosystem was smaller. But the psychology was identical. And the lesson, that a genuinely transformative technology does not guarantee that every company built on top of it will succeed, has had to be relearnt several times since.

    The BBC’s retrospective coverage of the dot-com collapse captures some of that atmosphere well, if you want a contemporary account of how British business press processed the wreckage. What it cannot quite convey is how total the belief had been. You had to be there, reading the Evening Standard in the back of a black cab heading east along the City Road, to understand just how completely rational people had convinced themselves that the old world was over.

    Frequently Asked Questions

    When exactly did the UK dot-com boom start and end?

    The UK dot-com boom gathered serious momentum from around 1997 but peaked in early 2000, when the NASDAQ hit its high point on 10 March 2000. The collapse played out through 2000 and 2001, with most British dot-com casualties folding before the end of 2001.

    Which UK companies were most famous for collapsing during the dot-com crash?

    Boo.com is the most cited example, burning through roughly £80 million before collapsing in May 2000. Clickmango.com, Deja.com (UK operations) and several others followed. Many smaller Shoreditch start-ups disappeared without significant press coverage.

    How much venture capital was invested in UK tech during the dot-com boom?

    British Venture Capital Association data from the period shows UK technology investment more than trebled between 1997 and 1999. Precise totals are hard to verify because many deals were structured across multiple European jurisdictions, but several billion pounds flowed into British internet companies during the peak years.

    Why did so many dot-com companies fail so quickly?

    Most were funded on the assumption that user growth alone would eventually translate into profit, a logic that collapsed once investors stopped providing fresh capital. Many also launched products that required broadband infrastructure to work properly, at a time when the majority of UK households were still on dial-up connections.

  • The History of UK Online Music Piracy: Napster, LimeWire and the Industry’s Long Fight Back

    The History of UK Online Music Piracy: Napster, LimeWire and the Industry’s Long Fight Back

    There is a particular kind of guilt that anyone who spent their teenage years in early-2000s Britain will recognise immediately. You had a dial-up connection, a stack of blank CD-Rs, and Napster open in the background. The download bar crept forward. The phone line crackled. You crossed your fingers that your parents would not pick up the receiver and kill it before the MP3 finished. Online music piracy UK history Napster stories belong to an entire generation, and I’d argue they shaped British music consumption more profoundly than anything since the cassette tape.

    Old desktop computer with blank CDs representing online music piracy UK history Napster era
    Photo by cottonbro studio on Pexels

    How Napster reached British teenagers before the record labels noticed

    Napster launched in the United States in June 1999, created by Shawn Fanning as a peer-to-peer file-sharing network. Within months it had spread to British universities, where students on JANET-connected networks could pull down albums in minutes rather than the hours dial-up required at home. By late 2000 it had roughly 70 million registered users worldwide, with a substantial British contingent. The British Phonographic Industry (BPI), which represents UK record labels, watched the numbers with mounting alarm. Physical single sales, which had been the engine of the UK charts for decades, began to dip in ways the industry found hard to explain away.

    The legal challenge to Napster came first from the American courts. Metallica’s lawsuit in April 2000 and the subsequent major-label cases forced Napster into a filtered, neutered version of itself and eventually into bankruptcy by 2002. For British teenagers, the reaction was practical rather than philosophical. You simply moved on to the next tool.

    The BPI’s legal campaigns and who actually got sued

    The Recording Industry Association of America filed lawsuits against individual American file-sharers from 2003 onwards, and the BPI followed with its own wave of legal action against UK users. In 2004 and 2005, the BPI sent letters to several hundred individuals believed to be sharing large quantities of music online, threatening civil action. The sums involved were sobering: the BPI argued that each illegally downloaded track represented lost revenue, and theoretical damages could run into thousands of pounds per person.

    In practice, the cases that actually reached court involved relatively small numbers of people, and the strategy was arguably more about deterrence than revenue recovery. The BPI secured a landmark ruling in 2008 requiring certain ISPs to hand over the details of users suspected of large-scale sharing. It was a sign of things to come: rather than pursuing millions of individual listeners, the industry began leaning on the infrastructure providers themselves.

    LimeWire had by this point replaced Napster as the tool of choice. It ran on the Gnutella network, which meant there was no central server to shut down. You installed the software, searched for a track, and connected directly to another user’s machine. It was faster and harder to kill than Napster. Kazaa, eDonkey, and eventually BitTorrent all played their part too, each iteration making the system more resilient. I remember the sheer implausibility of it all: that you could type an album title into a search box and it would simply appear, shared freely by strangers across the country and across the world.

    What the charts actually showed during the piracy years

    The UK singles chart had always been a reliable thermometer for the music industry’s health. From the mid-1990s through to the early 2000s, a successful single might shift 500,000 physical copies. By 2004, the numbers were dramatically lower. The Official Charts Company, which has tracked UK music sales since 1952, recorded consistent declines in physical single sales throughout the period. The BPI published annual figures showing year-on-year falls and attributed a significant portion directly to illegal downloading, though economists debated how much piracy displaced sales versus simply reflecting changing tastes.

    Albums held up somewhat better, particularly at the higher end of the market. But the mid-range artist who might once have counted on steady high-street sales found themselves particularly squeezed. Woolworths, which had been one of the largest music retailers in Britain, closed in 2008. HMV very nearly followed it. The physical infrastructure of the music industry was collapsing precisely as digital distribution took hold, and illegal distribution was moving far faster than legal alternatives.

    The move from burning CDs to streaming: a slow conversion

    iTunes launched in the UK in June 2004, offering tracks at 79p each. The BPI supported it enthusiastically, and it was genuinely a landmark moment: a legal, convenient, reasonably priced digital alternative had arrived. But the catalogue was incomplete, the tracks were initially DRM-locked (meaning you could not freely copy what you bought), and free alternatives remained a few clicks away. Adoption was gradual rather than sudden.

    Spotify launched its UK service in February 2009. The streaming model, free with adverts or paid for a monthly subscription, changed the calculus entirely. I think this is where the real turning point came, not in any courtroom. Streaming removed the friction. You did not have to wait for a download. You did not have to manage a library of files. You simply searched and played. By 2015, streaming had overtaken downloads as the primary form of digital music consumption in Britain, according to the BPI’s own data. By 2023, the BPI reported that streaming accounted for over 80 per cent of recorded music revenue in the UK.

    The piracy era did not end cleanly. The Pirate Bay, founded in Sweden in 2003, has been blocked by UK ISPs under court orders since 2012, though circumvention tools mean it remains accessible to determined users. But the scale of casual, habitual piracy that defined the Napster and LimeWire years has undeniably shrunk. When a streaming subscription costs less than a single CD once did, the calculation changes.

    What the piracy years actually changed in Britain

    The peer-to-peer era reshaped how British listeners related to music in ways that went beyond the legal and commercial. It democratised discovery. You could explore an entire back catalogue without risk. Obscure artists from the 1960s and underground acts from overseas found British audiences they would never have reached through traditional retail. Radio 1 and the high street had always been gatekeepers; Napster and its successors kicked the gate off its hinges.

    The dot-com crash of 2000 to 2001 had already rattled confidence in digital business models across Britain, as I’ve written about in the context of the British start-ups the City funded and the crash destroyed. The music industry faced its own version of that reckoning, just spread across a longer and more painful decade. And in the same way that the UK’s early internet culture was shaped by specific local factors, from the dial-up costs that made internet cafés so valuable to the way Freeserve democratised access, the piracy era had its own British texture. University networks, second-hand CD markets, and the particular culture of the UK music press all shaped how people encountered and shared music illegally.

    The BPI still publishes annual piracy reports and still works with ISPs to block infringing sites. But the organisation’s current focus has shifted substantially towards ensuring streaming platforms pay fairly to rights holders and artists, which is a rather different kind of fight. The battles of the Napster years feel, in retrospect, like a long and expensive lesson in what happens when an industry’s business model fails to keep pace with its audience’s behaviour. Britain’s listeners did not want to steal music. They wanted it to be easy and affordable. It just took the industry fifteen years to believe them.

    Frequently Asked Questions

    When did Napster become popular in the UK?

    Napster spread through British universities from around 2000, where fast academic network connections made it far more practical than dial-up at home. By 2001 it had millions of UK users, though the service was forced to shut down by legal action in 2002.

    Did the BPI actually sue individual music pirates in the UK?

    Yes. Between 2004 and 2005 the BPI sent legal letters to hundreds of British individuals suspected of large-scale file sharing, and a small number of cases resulted in civil settlements. The strategy was largely about deterrence rather than mass litigation, and later shifted to pressuring ISPs to block infringing sites.

    How much did illegal downloading affect the UK charts?

    Physical single sales dropped sharply from the early 2000s onwards, with chart-topping singles selling a fraction of what they had in the 1990s. The Official Charts Company and BPI attributed a significant portion of this decline to illegal downloading, though the rise of streaming eventually reversed the revenue trend.

    When did Spotify launch in the UK and how did it change things?

    Spotify launched its UK service in February 2009. By offering a free, ad-supported tier alongside a paid subscription, it made legal streaming more convenient than piracy for most casual listeners. The BPI’s data shows streaming overtook downloads as the primary format by around 2015.

  • Flash, Funnymouth and Newgrounds: How Browser Animations Took Over British School Computer Rooms

    Flash, Funnymouth and Newgrounds: How Browser Animations Took Over British School Computer Rooms

    There was a specific sound that preceded everything. The dial-up screech, the negotiation handshake, the brief silence, and then, if you were lucky, a tiny progress bar inching across a grey box while a Flash animation loaded one painful kilobyte at a time. I spent a remarkable portion of my secondary school lunch hours watching that progress bar. We all did. And when the thing finally played, if it was good, you’d call across the room to whoever was free. “Come and watch this.” The teacher on duty would glance over, see nothing obviously inappropriate, and go back to their marking.

    The Adobe Flash history UK web animations represent a very specific chapter of British internet culture, one that ran roughly from 1997 to around 2012, and left almost nothing behind. No archive. No museum. Just a faint collective memory and a plugin that Adobe finally killed off on 31 December 2020. What follows is an attempt to reconstruct that chapter before it fades entirely.

    Students in a school computer room during the Adobe Flash history UK web animations era
    Photo by Thành Đỗ on Pexels

    What Flash actually was, and why it mattered

    Flash began life as a product called FutureSplash Animator, created by a small company called FutureWave Software. Macromedia acquired it in 1996 and rebranded it as Macromedia Flash. Adobe bought Macromedia in 2005 and inherited the whole ecosystem. By that point, Flash was already everywhere. It ran on something like 98 per cent of desktop browsers worldwide, according to Adobe’s own figures at the time, and it had become the default delivery mechanism for anything on the web that moved, played music, or asked you to click something colourful.

    For British web users in the late 1990s and early 2000s, this mattered enormously because the alternative was almost nothing. Standard HTML at the time was static, tables-based, largely grey. Flash gave independent animators and game developers a way to publish work that actually looked and behaved like software. The barrier to entry was low enough that a determined teenager with a pirated copy of Flash MX and a dial-up connection could make something genuinely good, publish it to Newgrounds or a free Geocities page, and reach an audience of thousands.

    How animations spread across British schools and homes

    The distribution mechanism for early Flash content was almost entirely word of mouth, email forwarding, and a handful of aggregator sites. There was no YouTube. There was no social media in any recognisable form. If something was funny, you emailed the URL to everyone in your address book, and those people emailed it to theirs. In British schools, this meant the IT room became a kind of screening room. Someone would arrive having seen something at home the night before on a family dial-up connection, and they’d spend twenty minutes getting everyone else to load it before the bell rang.

    The animations that spread this way had a specific character. They were short, because short things loaded faster. They were loud and ridiculous, because a quiet understated piece did not survive the forwarding process. And they tended to be slightly transgressive in ways that felt edgy to a twelve-year-old but would barely register today. The legendary “Badger Badger Badger” by Jonti Picking (known online as Weebl) is the clearest British example: a looping animation of badgers accompanied by a maddening earworm that spread across UK schools in 2003 and genuinely seemed inescapable for months. Weebl was British, based in Cambridge, and his work had a distinctly UK flavour that you could sense even if you couldn’t quite articulate it.

    Newgrounds and the UK audience it found by accident

    Newgrounds was an American site, founded by Tom Fulp in Pennsylvania, but its audience was genuinely global and the British contingent was substantial. You can see this in archive discussions, in the old review threads, in the usernames that crop up on UK-specific forum posts from the period. British teenagers were heavy Newgrounds users throughout the early 2000s, partly because the site’s content moderation was loose enough to feel genuinely countercultural, and partly because there was simply very little else offering the same density of free interactive content.

    The specific animations that circulated most widely in British schools tended to be the ones that combined absurdist humour with surprisingly competent animation. “Salad Fingers” by David Firth, who is British, arrived in 2004 and spread through UK schools with the speed of something genuinely unsettling, which it was. Firth’s work had a homemade, almost lo-fi quality that felt entirely at home on a 15-inch CRT monitor in a draughty IT room. His animations were strange enough that teachers who caught students watching them could not quite decide whether to confiscate the URL. I’ve spoken to people who remember their IT teacher watching Salad Fingers with visible confusion rather than disciplinary intent.

    The email-forwarded link culture of this period has some parallels with how Friends Reunited spread through British workplaces and homes at roughly the same time, organic, word-of-mouth, carried by personal recommendation rather than any algorithm. People trusted what their mates sent them.

    The UK sites that hosted and spread Flash content

    Beyond Newgrounds, there were UK-specific hosting points for this material. Albino Black Sheep was Canadian but had heavy British traffic. b3ta.com, founded in London in 2001, became arguably the most important British hub for weird Flash animations and image macros. Rob Manuel’s site had a newsletter that circulated links to the best new material each week, and getting your animation featured in the b3ta newsletter was roughly equivalent to going viral by the standards of the time. The site still exists, which makes it an unusual survivor.

    Worth noting too is that the context in which British users encountered this material was shaped heavily by infrastructure. Broadband reached British homes far more slowly than many people now remember. As late as 2003, fewer than a quarter of UK internet connections were broadband, according to Ofcom figures. Most home users were still on dial-up, which meant loading a three-minute Flash animation at home was a commitment. You’d start it running and go and make a cup of tea. The school IT room, with its faster connection, was often where you had your first experience of something that would have taken twenty minutes to load at home. This created a particular social dynamic around watching, it was a communal activity in a way that streaming content today almost never is.

    The internet café culture of the same period offered another venue for this kind of communal viewing, though cafés were more associated with gaming than with animation. Still, if you were near a Cyberia or a local independent shop with a fast connection, you could load things there that would have taken forever at home.

    What’s actually been lost, and whether any of it survives

    The end of Flash as a functioning plugin in 2020 wiped out an enormous amount of content. Many animations that existed only on old hosting platforms, without source files or backups, simply stopped working. The Internet Archive’s Ruffle emulator project has preserved some material, Ruffle is an open-source Flash Player emulator written in Rust, and the Archive has used it to make thousands of old Flash files playable again through a web browser. But coverage is patchy. Anything that relied on server-side scripts, external assets, or ActionScript features that Ruffle hasn’t yet implemented simply doesn’t work.

    The British Library’s web archiving programme has captured some of this material, though its mandate is primarily focused on UK-registered domains. Much of the content British users actually consumed, Newgrounds, b3ta’s international traffic, the various forwarded files that lived on American servers, falls outside that scope. As I’ve noted when looking at other episodes of lost British internet history, the assumption that “it’s online so it will always be there” has proved catastrophically wrong for this era.

    The BBC has published some reflection on this period, and Ofcom’s historical data on UK internet adoption gives a partial picture of the infrastructure context. But there’s no dedicated British archive of early web animation culture. No equivalent of a film library for the things that made a generation of British teenagers laugh themselves breathless in school IT rooms between 1999 and 2008.

    My own memory of this period is probably not more reliable than anyone else’s. But I do think the Adobe Flash history UK web animations chapter deserves more serious attention than it has received. It was a genuine folk culture of the early web, distributed, unauthored in the traditional sense, consumed communally and then largely forgotten. The tools to study it are disappearing faster than the scholars to use them are appearing. That seems like something worth documenting while there are still people who remember the dial-up screech that preceded it all.

    Frequently Asked Questions

    What was Adobe Flash and when did it end?

    Adobe Flash was a multimedia software platform used to create animations, games, and interactive content for web browsers from the mid-1990s onwards. Adobe officially discontinued support for the Flash Player plugin on 31 December 2020, after which browsers blocked it from running entirely.

    What UK websites were known for hosting Flash animations?

    B3ta.com, founded in London in 2001, was probably the most significant British hub for Flash animations and weird web content. Newgrounds, though American, had a large British audience, and various free hosting platforms like Geocities were used by UK animators to publish their own work.

    Can old Flash animations still be watched anywhere?

    Some can. The Internet Archive (archive.org) has used an open-source emulator called Ruffle to make thousands of old Flash files playable in a modern browser. Coverage is incomplete, and anything relying on complex scripting or external servers may still not work correctly.

    Who made 'Badger Badger Badger' and where was it from?

    “Badger Badger Badger” was created by British animator Jonti Picking, who worked under the name Weebl. It was published in 2003 and spread rapidly through British schools and workplaces as one of the most recognisable Flash animations of the era.

    Why did Flash animations spread so quickly before social media existed?

    They spread primarily through email forwarding and aggregator sites. When someone found a funny or striking animation, they emailed the URL to contacts, who passed it on in turn. Sites like b3ta.com also published weekly newsletters curating the best new material, which acted as an early form of viral content distribution.

  • The Forgotten History of Ananova: The World’s First Digital News Presenter, Built in Leeds

    The Forgotten History of Ananova: The World’s First Digital News Presenter, Built in Leeds

    In the spring of 2000, a green-haired animated woman appeared on computer screens across Britain and was greeted as a genuine technological marvel. Her name was Ananova. She read the news, she moved her lips in something approximating speech, and she existed entirely as software. The Guardian, the BBC, and seemingly every technology correspondent in the country declared her the future of broadcasting. Eighteen months later, she had been sold to a mobile phone company, quietly repositioned as a novelty delivery mechanism for football scores, and more or less erased from public memory. This is what the Ananova virtual newsreader story tells us about the dot-com era: that novelty and sustainability were, all too often, two entirely different things.

    Early 2000s newsroom computer screen recalling the era of the Ananova virtual newsreader
    Photo by Ruben Boekeloo on Pexels

    Who built Ananova, and why

    The Press Association, Britain’s national news agency, launched a digital division in the late 1990s under the banner PA New Media. The brief was straightforward enough for the time: find a way to deliver journalism on the internet. What the team in Leeds actually built went considerably further than a news wire with a website attached. They created a synthetic presenter, a computer-generated figure who could receive a text script, render it as speech, and animate a face around that audio in something close to real time. The technical challenge was significant. This was 2000, remember, not 2020. Rendering a talking human face required real engineering effort, and the Leeds team did it without the GPU horsepower that would later make such things routine.

    The character herself was designed with a deliberate aesthetic. The green hair was a choice, not a glitch. PA New Media wanted Ananova to read as futuristic rather than as a failed attempt at human realism. She had a name, a face, a measured mid-Atlantic accent, and a slot at the top of the PA website that served actual breaking news. When she announced the result of the 2000 US presidential election recount, several publications ran the story with the angle that a virtual presenter had beaten their human counterparts to air. That detail, true or not in its particulars, captured the mood of the moment.

    The £95 million sale and what happened next

    In May 2000, at the absolute peak of dot-com valuations, Orange purchased Ananova from PA for £95 million. The figure was extraordinary for a piece of software and a brand that had existed for barely a few months. I’ve looked back at the coverage from that period and the sheer confidence in the reporting is striking. Orange was paying not just for what Ananova was, but for what commentators assumed she would become: an always-on, personalised digital news anchor who would sit at the centre of internet-delivered media.

    Orange’s actual use for Ananova turned out to be narrower. The virtual newsreader element was gradually wound down. The animated face that had launched with such fanfare became, in practice, expensive to run and difficult to scale on the mobile networks of the early 2000s. What survived was the Ananova brand attached to a text-based news and sport aggregation service, which Orange ran for several years before that too was folded into its other digital properties. The Leeds team dispersed. The original technology was not, as far as any public record indicates, developed further by its new owners.

    Vintage dial-up modem representing the internet infrastructure limitations that constrained the Ananova virtual newsreader
    Photo by Petra Ryan on Pexels

    Why the technology was ahead of its infrastructure

    This is where the Ananova story becomes genuinely instructive rather than simply sad. The virtual newsreader concept was not wrong. Synthetic media presenters are now entirely routine: weather forecasts, corporate training videos, multilingual news services. What Ananova lacked was the surrounding infrastructure that would have made her practical at scale. Broadband penetration in Britain in 2000 was tiny. The broadband race was barely underway, BT’s ADSL rollout was still grinding through its early phases, and most people accessing the Ananova website were doing so over a dial-up connection. Streaming a real-time animated talking head over a 56k modem was, to put it charitably, not a smooth experience.

    The mobile network argument Orange made was similarly premature. 3G services in Britain were years away from meaningful coverage, and the handsets of 2001 and 2002 could not meaningfully render what Ananova’s face required. So the technology sat, waiting for infrastructure that would not arrive until it was too late to matter. By the time broadband was genuinely widespread and mobile video was practical, synthetic media had been rebuilt from scratch by other companies with better funding and no legacy baggage.

    This pattern, of British internet technology being genuinely innovative and then being sold or stranded before it could benefit from the conditions it had anticipated, recurs throughout the dot-com period. The dot-com crash took down dozens of British ventures that had legitimate ideas but insufficient runway. Ananova was slightly different in that she was sold rather than collapsed, but the outcome for the technology was similar.

    What Ananova meant for British internet culture

    I’d argue that Ananova mattered beyond her own brief lifespan for a specific reason: she was a moment when the British public, not just the technology industry, took seriously the idea that the internet could replace rather than merely supplement traditional media. She appeared on television news programmes. She was profiled in national newspapers. She generated genuine public curiosity rather than the niche enthusiasm that surrounded most internet developments of the period.

    That curiosity reflected something real about the moment. People had been getting online in growing numbers through services like Freeserve, visiting internet cafés on the high street, and beginning to understand that this technology was not going away. Ananova arrived at precisely the point when the question “can the internet do journalism?” felt urgent and genuinely open. She offered an answer that was theatrical and memorable, even if it turned out to be premature.

    The PA New Media team were also, it should be said, doing something culturally distinct from what was happening in Silicon Valley. This was not a start-up founded on venture capital and optimism. It was a journalism organisation, one of the oldest and most reliable in Britain, attempting to imagine its own future. That context matters. The Ananova project was an act of institutional imagination, and those are rarer and more interesting than they sometimes appear.

    What the record shows now

    Finding reliable documentation of Ananova’s technical architecture is not straightforward. Much of what was written about her in 2000 was promotional rather than analytical, and the PA New Media team did not publish detailed technical papers. The BBC’s technology coverage of the period gives some sense of the contemporary reception, and the Wayback Machine preserves fragments of the original Ananova site, though the animated elements no longer function. This is a recurring problem with early internet history: the things that made a technology remarkable are often precisely the things that do not survive archiving.

    Groups working to understand and document the digital economy’s early years, from academic researchers to organisations like Activism Digital who track how internet culture intersects with public communication, often cite Ananova as an early instance of synthetic media entering mainstream consciousness. The virtual newsreader was not a footnote. She was a signal, one that the market misread and the infrastructure could not yet support, but a genuine signal nonetheless.

    What I find most striking, looking back at the Ananova experiment, is how completely she was forgotten despite how loudly she was celebrated. The £95 million sale was front-page news. The wind-down was barely a footnote. That asymmetry, between the noise of launch and the silence of discontinuation, is perhaps the most honest summary of what the dot-com era was like from the inside. The future arrived, briefly, in Leeds in 2000. Then it was sold, and mostly went quiet.

    Frequently Asked Questions

    What was Ananova and when was she launched?

    Ananova was the world’s first virtual newsreader, a computer-generated animated presenter built by PA New Media in Leeds. She launched publicly in April 2000 and could read breaking news scripts aloud using synthetic speech and facial animation.

    How much did Orange pay for Ananova?

    Orange purchased Ananova from the Press Association in May 2000 for £95 million, one of the larger acquisitions of the dot-com boom period. The sale reflected the extraordinary valuations placed on internet technology at that moment, rather than Ananova’s actual revenue.

    Why did Ananova fail as a virtual news presenter?

    The technology outpaced the available infrastructure. Streaming an animated talking head in 2000 required broadband that most British homes did not yet have, and mobile networks were too slow to carry video effectively. Orange shifted Ananova toward text-based content delivery and the original animated presenter concept was quietly dropped.

    Is Ananova still around in any form?

    The Ananova brand survived in a much reduced form as a mobile content and news aggregation service under Orange for several years after the acquisition, but it was eventually wound down. The original animated newsreader technology was not developed further by Orange and has not been revived.

  • How Friends Reunited Gave Britain Its First Taste of Social Networking, and Then Lost Everything

    How Friends Reunited Gave Britain Its First Taste of Social Networking, and Then Lost Everything

    There is a particular kind of digital ghost story that the early 2000s produced more than any other era of computing. A platform appears, seemingly from nowhere, captures an entire nation’s attention, earns a small fortune, and then evaporates, leaving behind only the memories of people who once spent their lunch breaks searching for old school friends they half-remembered from the third form. Friends Reunited history is exactly that kind of story. And it is, I think, one of the most poignant chapters in British internet history precisely because it felt so genuinely British: cautious, slightly nostalgic, and ultimately overtaken by something louder from across the Atlantic.

    Early 2000s home computer setup evoking the era of Friends Reunited history and Britain's first social networks
    Photo by Evelyn Rumti on Pexels

    The website launched in July 2000, built by a husband and wife team, Steve and Julie Pankhurst, working out of their home in Barnet, north London. The premise was arrestingly simple. You registered, you found your old school, you listed yourself as a former pupil, and you could see who else had done the same. That was essentially the entire product. No algorithmic feed, no sponsored posts, no elaborate privacy settings. Just a directory of former classmates and a small text box where you could write what you had been doing with your life since 1987.

    The Pankhursts built the initial version themselves, with technical help from Jason Porter, a friend of Steve’s. Early traffic was modest. In the first few months, the site attracted a few thousand registrations, encouraging, but hardly the kind of numbers that would make anyone very excited. Then, in the spring of 2001, something shifted. Word of mouth, spread largely through office email chains and the nascent habit of forwarding links to colleagues, sent registration numbers into a steep upward curve. By the end of 2001, the site had around 4.5 million registered members. By 2003, that figure had reached 15 million, in a country of roughly 60 million people, that is a remarkable penetration rate for any web service, let alone one that charged a subscription fee.

    The subscription model that should have doomed it but didn’t

    This is the detail that always surprises people when I recount the Friends Reunited story. The site charged users £7.50 per year to send messages to former classmates. Free users could browse and be seen, but they could not initiate contact. By the standards of today’s social media, where every platform operates on the assumption that users will never pay directly for access, this seems bizarre. In 2001 and 2002, it worked. The hunger to reconnect with people from school, from old jobs, from university, was strong enough that millions of British people happily handed over their credit card details. At its peak, Friends Reunited was generating around £2 million per month from subscriptions alone, not counting the advertising revenue that came as traffic grew.

    Part of what made the site work was its emotional proposition. The internet of 2001 was still, for many British users, primarily a tool for searching, reading the news, and sending email. The idea that it could facilitate genuine human reconnection, that you might, with a few clicks, locate someone you had not spoken to since your GCSEs, felt almost magical. I remember conversations from that period where people described finding old friends on the site with a kind of hushed excitement that you simply do not associate with opening Instagram today. The novelty was total.

    Old school photographs on a table, representing the nostalgia at the heart of Friends Reunited history
    Photo by Wolfgang Vrede on Pexels

    Why ITV paid £120 million for it

    By late 2005, Friends Reunited was attracting serious acquisition interest. ITV, then struggling to understand what the internet would eventually do to broadcast television, acquired the site in December 2005 for £120 million. From a broadcaster’s perspective, the logic was reasonable, if perhaps slightly panicked. The site had tens of millions of registered British users. It had a subscription revenue stream. It had enormous brand recognition. For a company trying to build a digital presence quickly, that looked like a shortcut worth taking.

    What ITV perhaps did not fully account for was the speed at which the market was moving. Facebook had launched in February 2004, initially restricted to American university students. By late 2006, it had opened to anyone with an email address. MySpace was already enormous. Bebo, which briefly dominated among British teenagers in a way that is worth its own chapter (and which I’ve written about separately, the story of Bebo’s conquest of British teenagers is a genuinely fascinating parallel tale), was pulling users away from more traditional connection-based platforms. The competitive landscape that Friends Reunited had briefly occupied alone was, by 2007, extraordinarily crowded.

    ITV’s stewardship was not, by most accounts, inspired. The subscription fee was eventually dropped, which was probably the right decision, but the site struggled to find a clear identity in a world where Facebook offered everything it did and considerably more, for free. The advertising model that sustained other social platforms required genuine daily engagement, the kind of habitual, reflexive checking that Facebook was engineering with considerable sophistication. Friends Reunited had been built around a one-time emotional impulse: find your old classmates. Once you had found them, the reason to return evaporated. Publishers and brands trying to reach audiences on the platform through Banner Ads found diminishing returns as traffic thinned out through 2008 and 2009.

    The collapse and what it actually tells us

    ITV sold Friends Reunited in 2009 to DC Thomson, the Dundee-based publisher behind The Dandy and The Beano, for just £25 million, a loss of £95 million in four years. DC Thomson attempted a relaunch, reframing the site as a platform for sharing old photographs and memories rather than active social networking. It was a sensible repositioning, leaning into what Friends Reunited had always done better than Facebook: the specific, slightly melancholy pleasure of looking backwards. But the audience did not follow in sufficient numbers. The site was shut down permanently in February 2016.

    The story of Friends Reunited sits alongside other cautionary tales of the British dot-com era, the frantic funding, the rapid acquisition, the collapse, which I’ve examined at length in the context of the broader UK dot-com casualties of that period. But Friends Reunited is slightly different from the companies that were simply overfunded and underconceived. The Pankhursts built something that genuinely worked, that solved a real emotional need, and that millions of people actually used. The failure was not in the original idea. It was in the inability, shared, to be fair, by almost everyone in 2005, to anticipate how completely Facebook would redraw the rules of online social interaction.

    There is a certain irony in the fact that Friends Reunited was acquired for £120 million at almost exactly the moment it began to decline. The rise and fall of Friends Reunited is a story about timing as much as anything else. The Pankhursts built the right product for the exact right moment in British internet history. The site captured something real about how people in this country felt about connection and memory. That is not nothing. It is, I would argue, considerably more than many of the platforms that outlasted it ever managed.

    According to the BBC’s coverage of the closure in 2016, Steve Pankhurst reflected that the site had simply been overtaken by the scale of what followed. That is an understatement, but it is also, in its quiet British way, exactly the right assessment. Friends Reunited was first. Being first, as British internet history keeps demonstrating, is rarely enough on its own.

    Frequently Asked Questions

    Who founded Friends Reunited and when was it launched?

    Friends Reunited was founded by Steve and Julie Pankhurst, with technical assistance from Jason Porter, and launched in July 2000. The couple built the site at their home in Barnet, north London, initially as a way to help people reconnect with former classmates.

    How much did ITV pay for Friends Reunited?

    ITV acquired Friends Reunited in December 2005 for £120 million. The broadcaster hoped the site’s large registered user base would accelerate its digital strategy, but sold it on to DC Thomson in 2009 for just £25 million, a loss of £95 million in under four years.

    Why did Friends Reunited fail?

    The site’s decline was driven primarily by the rapid rise of Facebook, which offered free messaging, a dynamic news feed, and daily engagement features that Friends Reunited could not match. Once users had reconnected with old classmates, the site’s core appeal, there was little reason to return regularly.

  • The CD-ROM That Changed British Internet History: How Dixons and Freeserve Started a Price War Nobody Saw Coming

    The CD-ROM That Changed British Internet History: How Dixons and Freeserve Started a Price War Nobody Saw Coming

    There is a particular kind of historical accident that only becomes legible in retrospect. In the summer of 1998, Dixons, the high street electronics retailer best known for flogging video recorders and washing machines, bundled a compact disc into a cardboard sleeve, stacked those sleeves near the tills of its shops, and quietly changed the shape of British domestic life. The disc was for Freeserve. And almost nobody inside Dixons expected it to work quite the way it did.

    A CD-ROM disc in a cardboard sleeve representing the Freeserve free ISP distribution method
    Photo by Arturo A on Pexels

    I’ve spent a good deal of time reading through the contemporary trade press coverage of this period, and what strikes me most is how little fanfare accompanied the launch. The Financial Times treated it as a minor retail curiosity. The specialist computing press was mildly sceptical. The idea itself, after all, was not especially original: bundle an ISP onto a CD-ROM, give it away free, collect the revenue share from the phone calls. America Online had been doing something broadly similar in the United States for years, showering the postal system with unsolicited discs. What made Freeserve genuinely different, and genuinely significant, was the cost model underneath it.

    What Freeserve actually changed about getting online in Britain

    Before Freeserve arrived, getting connected to the internet in Britain required writing a cheque. Services like Demon Internet, which had done so much to build a culture of early British net use, charged a monthly subscription fee. AOL, which had launched in the UK in 1996, charged similarly. You paid for access, and then you paid again every time you picked up the phone, because dial-up connections ran over standard telephone lines billed by the minute. It was not cheap. For a household on an ordinary income, the internet in 1997 was still a luxury with a running cost to match.

    Freeserve eliminated the subscription. There was no monthly fee. You loaded the disc, you signed up, and you were online. Dixons made its money not from the customer but from the telephone companies: every minute a Freeserve user spent online generated a small payment back to the ISP via a revenue-sharing arrangement with BT. The customer paid only the ordinary cost of the phone call, the same rate they would pay to ring a local number. That was the insight. And it worked because British telecoms regulation at the time made that revenue share model viable in a way it was not in every other country.

    The Dixons connection: why a retailer built an ISP

    The man usually credited with the idea is Ajaz Ahmed, who pitched it to Dixons’ then chief executive, John Clare. The story, as it has been retold many times, is that Ahmed proposed the concept in a meeting that lasted roughly twenty minutes, was given a budget of around £500,000, and had Freeserve operational within about a year. Whether every detail of that narrative is precise I cannot say with certainty, but the broad outline holds up. Dixons was looking for a way to sell more computers. If you could demonstrate to a potential buyer in the shop that internet access was free, that removed one of the biggest objections to buying a home PC in 1998. The ISP was, in a sense, a loss-leader for the hardware business.

    What nobody anticipated was the velocity of the uptake. Within six months of launching in September 1998, Freeserve had more registered users than any other ISP in the country. By early 1999, the figure was over a million. To put that in context, it had taken Demon Internet the better part of five years to reach a fraction of that number. Freeserve had not simply entered the market; it had restructured it. You can read more about what the free ISP revolution actually meant for British internet history in a broader treatment of the topic, but the short version is this: after Freeserve, subscription-based dial-up access was finished as a mass-market proposition.

    The price war that followed

    The competitors noticed immediately. LineOne, which had the backing of BT and News International, dropped its fees. AOL UK introduced a free tier. Dozens of smaller ISPs followed the same revenue-share model. By mid-1999, the UK had well over a hundred free ISPs. The BBC reported at the time that the number of British households online was rising faster than almost any comparable European nation. The CD-ROM had become an unexpected instrument of mass access.

    There is a minor irony worth noting here. The very proliferation that Freeserve triggered eventually made the market extremely crowded and thin on margins. Revenue share from telephone calls was never enormous; spread across hundreds of competitors, it became vanishingly small. Many of those free ISPs collapsed in the dot-com crash of 2000 and 2001, as the wave of British start-up failures took down businesses that had been running on speculative valuations rather than sustainable income. Freeserve itself floated on the London Stock Exchange in July 1999 at a valuation that, even then, raised eyebrows. The shares surged. Analysts projected extraordinary growth. The arithmetic of the underlying business was less convincing.

    How Freeserve compared to AOL and what it meant for ordinary users

    AOL’s approach to the internet was, to put it charitably, proprietary. British AOL users in the late 1990s spent much of their time inside AOL’s own walled environment: AOL-specific chat rooms, AOL-specific content pages, AOL-specific email. The open web was accessible, but the design philosophy pointed you inward. Demon Internet, by contrast, was very much for people who already knew what they were doing. Its culture was technical, even a little austere. It had helped to shape the early internet in Britain, but it was not building mass adoption.

    Freeserve pointed users directly at the open web from the start. The portal it launched alongside the ISP service was basic, certainly, but it did not try to enclose its users. In that sense, it was closer to what we would now recognise as the model of the modern ISP: get people connected, let them go where they like. For many British families, Freeserve was their first experience of the internet in any form. The internet café had offered a taste, but Freeserve brought the thing home.

    The Wanadoo acquisition and what it tells us about internet consolidation

    Dixons sold Freeserve to the French telecommunications company Wanadoo, a subsidiary of France Télécom, in 2001. The sale price was around £1.65 billion. For context, this was still a substantial sum even after the dot-com crash had stripped away the more fantastic valuations. Wanadoo absorbed Freeserve’s user base and rebranded the service, first under the Wanadoo name and then, after France Télécom restructured, under the Orange brand. The Freeserve name disappeared. The brand that had been, for several years, the most recognised ISP in the country was gone.

    This pattern of acquisition and rebrand is one I find quietly instructive. The services that first brought mass audiences online rarely survived as independent entities. They were valuable precisely because they had users, and once those users existed, the question of who actually owned the infrastructure to serve them became commercially urgent for the telecoms giants. Freeserve’s story is in some ways the story of the entire early commercial internet: built fast, scaled unexpectedly, acquired, absorbed, renamed, forgotten. Companies like Inuvate represent a very different generation of digital business, one that emerged after those early convulsions had resolved into something more stable, but it is worth remembering how chaotic and contingent the founding era actually was.

    Why the Freeserve story still matters

    The question I keep coming back to is whether Freeserve was genuinely transformative or whether it simply arrived at the right moment. My honest reading is that it was both. The revenue-share model had been available to anyone willing to use it; Dixons used it because it had a distribution network of high street shops and a commercial incentive to sell PCs. The timing mattered because British households were, by 1998, broadly ready to consider getting online if the cost barrier came down. Freeserve removed the barrier. What happened next, the millions of new users, the price war, the flotation, the eventual absorption into a French telecoms empire, flowed from that one act of accidental disruption.

    The CD-ROM itself, that slim plastic disc in its cardboard sleeve, is now a relic. You would have to look quite hard to find one in working condition. But the change it carried is still with us, in every broadband-connected home in the country, in the assumption that getting online is simply what you do, rather than a specialist hobby with a monthly invoice attached. That assumption began, more than anywhere else, with a pile of discs stacked next to the tills of a Dixons on the high street. History rarely announces itself so quietly.

    Frequently Asked Questions

    What was Freeserve and when did it launch?

    Freeserve was a British internet service provider launched by the retailer Dixons in September 1998. It offered dial-up internet access with no monthly subscription fee, making it the first major free ISP in the UK and rapidly becoming the country’s largest by user numbers.

    How did Freeserve make money if it was free?

    Freeserve used a revenue-share arrangement with BT. Every minute a user spent online generated a small payment back to Freeserve from the telephone call revenue, without any extra charge to the customer beyond their normal local call rate. This model was made possible by the structure of UK telecoms regulation at the time.

    How was Freeserve different from AOL in Britain?

    AOL built a largely enclosed environment, steering users towards its own content, chat rooms, and services within a proprietary walled garden. Freeserve directed users to the open web from the start and charged no monthly subscription, making it a more open and accessible option for first-time internet users in British households.

  • Ceefax vs the Web: How the BBC’s Teletext Service Held On Long After the Internet Should Have Killed It

    Ceefax vs the Web: How the BBC’s Teletext Service Held On Long After the Internet Should Have Killed It

    There is a particular kind of stubborn loyalty that the British have for things that simply work. Ceefax, the BBC’s teletext service, ran from 1974 until 16 October 2012, and for roughly fifteen of those years it was sharing the nation’s attention with the World Wide Web. By all rights the internet should have made it irrelevant almost immediately. Faster, richer, searchable, illustrated: the early web looked, on paper, like everything Ceefax was not. And yet millions of people kept pressing the Text button on their remote controls, long into the broadband era. I find that fact genuinely fascinating, and I think it tells us something honest about how new technology actually replaces old technology, which is slowly, unevenly, and often not until the infrastructure is physically dismantled.

    A CRT television displaying a teletext page, evoking the Ceefax BBC teletext history end era
    Photo by Anete Lusina on Pexels

    What Ceefax actually was, and why it mattered

    Ceefax launched on 23 September 1974 as the world’s first public teletext service. It worked by encoding data into the unused lines of the television signal, a part of the broadcast that had previously carried nothing. No telephone line required, no subscription, no modem warming up on a desk. Every television in the country that could receive BBC1 or BBC2 was already receiving Ceefax, whether its owner knew it or not. The only cost was the set-top decoder chip, which became standard in new televisions through the late 1970s and 1980s. By the time the internet became a household word, Ceefax had already been part of British domestic life for two decades.

    The content was arranged across numbered pages. Page 101 was the index; 301 was football results; 302 was cricket; 400 was weather; 570 onward covered subtitles. Regular users had these page numbers memorised the way later generations would memorise URLs. The BBC published audience research in the early 2000s suggesting around twenty million people a week were still consulting the service, numbers that would have embarrassed many websites of the time. Part of that loyalty was habit. A larger part was practicality.

    What early websites genuinely could not do

    I have spent a long time reading through contemporaneous accounts of early internet use, and the honest assessment is that the web circa 1997-2002 was, for most British homes, a frustrating experience. The broadband rollout was patchy and slow, and for the majority of households in the late 1990s, getting online meant dialling up at 56 Kbps, tying up the phone line, and waiting. A Ceefax page loaded in under a second. It was always on.

    Football scores are the clearest example. On a Saturday afternoon in 1999, a fan who wanted live scores had three realistic options: radio, Ceefax page 302, or the web. The radio gave you one game at a time. The web, if you could get a dial-up connection running, served pages from BBC Sport or Sky Sports that were text-heavy but took thirty or forty seconds to load and required active refreshing. Ceefax cycled through all its pages automatically, refreshing every few minutes. You pressed 302, left it on the screen, and glanced up. That interaction model, passive, ambient, television-native, was something the early web simply could not replicate.

    Weather forecasts worked the same way. The Met Office had a web presence early, but the Ceefax weather pages gave five-day regional forecasts in a form that appeared instantly on a screen already open in the living room. No browser, no connection, no waiting. For older users in particular, and for households that did not yet own a home computer, this was not a minor convenience. It was the entire equation.

    The BBC’s own internal conflict

    What makes the Ceefax story more complex is that the BBC was simultaneously running Ceefax and building the website that would theoretically replace it. The BBC’s online presence launched in 1997, and by the early 2000s bbc.co.uk had become one of the most visited domains in Britain. The corporation was effectively competing with itself across two channels. Resources went to both. Ceefax editors kept writing and updating pages; web editors kept building and improving the site.

    There were internal arguments about priorities, and occasional memos from senior figures questioning why the BBC was maintaining what amounted to a 1970s data transmission system alongside a modern website. The counter-argument, which held for years, was the audience numbers. Twenty million weekly users is not a rounding error. The BBC could not simply switch Ceefax off without alienating a substantial and loyal portion of its licence fee payers. The BBC website was growing fast, but it had not yet absorbed those users.

    Digital switchover and the real reason for the 2012 end date

    Ceefax did not die because audiences abandoned it. It died because the analogue television signal was switched off. The UK’s digital switchover programme, coordinated region by region between 2008 and 2012, removed the analogue broadcast infrastructure that Ceefax had always depended on. Freeview, the digital terrestrial replacement, carried its own text service called Freeview text or MHEG-based interactive services, but these were not Ceefax. The BBC’s decision to keep Ceefax running on the remaining analogue transmitters right until the end of the switchover programme meant the final pages were transmitted from the Crystal Palace transmitter at 23:32 on 16 October 2012, when the London analogue signal went dark.

    The BBC marked the occasion. Page 100, the main index, was replaced with a simple farewell message. Screenshots circulated online within hours, and a generation of British viewers who had grown up with the service wrote about it with a warmth that surprised some commentators. The BBC’s own coverage of the switchoff ran longer than anyone had expected, precisely because reader response was so unexpectedly emotional.

    I think the emotion came from recognition. Ceefax had not just been a useful tool. It had been a texture of British life, the bouncing ball graphic on the loading pages, the chunky pixel font, the particular shade of yellow used for sport headlines, the satisfying click of typing page numbers on a remote control. These things accumulate meaning over thirty-eight years.

    What the Ceefax era actually tells us about digital transition

    The received wisdom about technological change is that new platforms eliminate old ones cleanly. In practice, the record is messier. Ceefax persisted because it had genuine advantages, always-on, no connection required, television-native, that the early web took years to match. Broadband penetration in Britain only crossed fifty per cent of households around 2007, according to Ofcom data. For the decade between the web’s arrival and ubiquitous fast connectivity, Ceefax was not a relic. It was often the better tool for specific tasks.

    The same pattern played out elsewhere. You can trace it in the broader history of teletext in Britain or in the way services like Friends Reunited held audiences even after Facebook launched, simply because they were familiar and required no learning curve. Technology adoption follows infrastructure, not announcement dates. Ceefax lasted as long as the analogue signal lasted. No longer, and not much shorter.

    The page numbering system, the chunky graphics, the automatic cycling: none of it could survive in a broadband world on its technical merits alone. But for its audience, Ceefax had earned the right to a long twilight. Thirty-eight years of reliable, instant information, free at the point of use, on a screen already sitting in the corner of the room. That is a harder act to follow than most people remember.

The Latest