4. Web 1.0: The Read Era (1990s–early 2000s)

The preceding chapter traced how the Internet and the World Wide Web came into being: a packet-switched network of networks, and, layered on top of it, Tim Berners-Lee's system of linked documents addressed by URLs, transmitted by HTTP, and written in HTML (see Chapter 3, The Internet and the World Wide Web: Origins and Architecture). This chapter examines what that early Web was actually like to use and to build, and why its particular character matters to the central question of this book. The argument is that the first decade or so of the public Web — what is now called "Web 1.0" — was overwhelmingly a read medium: a vast, growing library that most people consumed and few people contributed to. Understanding why publishing was hard, and how the Web's open and federated design coexisted with this lopsided reality, gives us the baseline against which the later "read–write–own" framing of Web1, Web2, and Web3 is defined. Without a clear picture of the "read" era, the claims made for the eras that follow cannot be properly assessed.

A term invented after the fact

It is important to begin with a point about the term itself, because it shapes everything that follows. Nobody in 1995 believed they were using "Web 1.0." The label is a retronym — a new name coined for an old thing once a successor makes the old thing look like merely the first version of something.

In plain terms: a retronym is a word we invent backward in time. "Acoustic guitar" was just "guitar" until the electric guitar arrived and we needed a way to distinguish the two. Likewise, the early Web only became "Web 1.0" once people started talking about "Web 2.0."

This sequence is the reverse of how it intuitively seems. The phrase "Web 2.0" appeared first, in 1999, in an article by the information designer Darcy DiNucci titled "Fragmented Future" (DiNucci, 1999); it was then popularized in 2004 when O'Reilly Media and MediaLive International launched the Web 2.0 Conference (Wikipedia: Web 2.0, n.d.; O'Reilly, 2005). Only against that "2.0" did the preceding period acquire the name "Web 1.0." The chapter, in other words, describes a historical period through a category that the period's own inhabitants did not possess. This is not a pedantic aside. The whole "read / read–write / read–write–own" schema that the Web3 movement uses to tell its own story (examined directly in Chapter 25, Toward a Definition: "Read, Write, Own" and Competing Framings) is a retrospective narrative — a tidy three-act drama imposed on a messier history. Recognising that the periodization is a later interpretive device, not a contemporaneous fact, is part of reading the Web3 thesis critically rather than swallowing it whole.

With that caveat in place, we can ask what the "read era" actually consisted of.

What the early Web was actually like

The Web became a public service on the Internet on August 6, 1991, and CERN placed its underlying software in the public domain in 1993, removing any licensing barrier to its adoption (CERN, n.d.). The very first web page was, fittingly, an instructional document explaining what the Web was and how to use it (Techopedia, n.d.). From that seed the medium grew, but its essential texture for most of the next decade was remarkably consistent: it was a collection of static pages.

A static page is, in essence, a printed poster pinned to a wall: every visitor sees exactly the same thing, and the only way the poster changes is if the author climbs up and replaces it by hand. A static web page is a fixed document — text, images, and links — that the server delivers identically to everyone, with no part of it generated freshly for the individual reader.

This contrasts with the dynamic page that would later dominate, in which the server assembles a custom document on the fly for each visitor — your Facebook feed, your Amazon recommendations, your search results. In the Web 1.0 period such dynamic, personalized assembly was rare, expensive, and technically demanding; the norm was the hand-built document that sat unchanged until its author edited the file and uploaded it again.

The genres of the early Web followed from this. There were personal homepages — idiosyncratic, hand-coded pages where individuals posted hobbies, photographs, lists of favourite links, and the era's characteristic "Under Construction" graphics. There were institutional and corporate "brochure" sites, which did online roughly what a printed company pamphlet did on paper: state who we are, what we sell, and how to reach us. And because the Web was growing faster than anyone could mentally map, there were two kinds of finding aids.

The first was the web directory: a human-curated, hierarchical catalogue of sites organized by topic, like a library's subject classification. Yahoo!, founded in 1994, began life precisely as such a directory — a list of interesting links that two Stanford graduate students maintained and sorted into categories by hand. The second was the portal: a single "front door" page that aggregated directories, news headlines, weather, email, and search, hoping to become the place a user started every browsing session. Early portals such as Yahoo!, Excite, Lycos, and AltaVista competed to be that default starting point.

In plain terms: imagine the Web as a fast-growing city with no street signs. A directory was a printed phone book that someone compiled by walking the streets and writing down every shop by category. A portal was the grand central station everyone passed through on the way in. Both existed because there was, as yet, no good automatic way to find anything — search engines were primitive, and Google (founded 1998) would not yet dominate for several years.

The most emblematic single service of the period was GeoCities, founded in November 1994 by David Bohnett and John Rezner. GeoCities let ordinary users build free personal homepages, and it organized them into themed "neighborhoods" — pages about technology went in one district, pages about entertainment in another, as though the homeless documents of the Web were being given postal addresses in a virtual town (Wikipedia: GeoCities, n.d.). The model resonated: by December 1995 GeoCities had grown to around fourteen neighborhoods and was serving more than six million page views a month, and by the time Yahoo! acquired it on January 28, 1999, it was reportedly the third-most-visited site on the entire Web (Wikipedia: GeoCities, n.d.).

GeoCities matters to this chapter for a reason beyond nostalgia. It is the strongest counter-example to a crude reading of "Web 1.0 was read-only" — here were millions of ordinary people publishing. At least roughly thirty-eight million user-built pages are estimated to have existed before the service was finally shut down in 2009 (Wikipedia: GeoCities, n.d.). The honest version of the "read era" thesis, then, is not that publishing was impossible but that it was hard, marginal, and unevenly distributed — and that the friction shaped the medium's social structure. To that friction we now turn.

The "read-only" character: many readers, few publishers

The defining structural fact of Web 1.0 is an asymmetry. A relatively small number of people and organizations produced content; a very large number consumed it. Web 1.0 is conventionally characterized as a period of largely static, read-only pages offering little user interaction beyond reading and following hyperlinks (Techopedia, n.d.; Wikipedia: Web 2.0, n.d.). When Tim O'Reilly later articulated what made "Web 2.0" different, the contrast he drew was precisely with this top-down publishing model: the new pattern, he argued, would harness collective intelligence through "an architecture of participation," turning users from an audience into contributors (O'Reilly, 2005). The implication is clear — the prior architecture was one of publication to an audience rather than participation by it.

It is worth being precise about what "read-only" does and does not mean here, because the phrase is a slogan that flattens a more interesting reality.

It does not mean the early Web was literally incapable of accepting input. From very early on there were ways for users to write back: email links, guestbooks where visitors could leave a comment, web forms, message boards and bulletin-board systems, Usenet newsgroups, and Internet Relay Chat. People conversed online throughout the 1990s. The point is not the absolute absence of two-way communication but its relationship to the core publishing layer — the web pages themselves. Those pages were, overwhelmingly, authored by their owners and merely viewed by everyone else. Writing to the Web — creating a page that others could find and read — was a categorically different and harder act than reading from it.

In plain terms: in the Web 1.0 world, reading was like watching television and writing was like building your own television station. A guestbook let you phone in to a station, but it did not make you a broadcaster. The asymmetry that defines the "read era" is exactly this gap between the ease of tuning in and the difficulty of going on air.

Why was going on air so hard? The answer is the crux of the chapter, and it has three intertwined components: technical skill, infrastructure, and cost.

Why publishing was hard

You had to be, in effect, a small-scale engineer

To put a page on the Web in 1996, a person typically had to do several things that today are invisible or automated. You had to write HTML — HyperText Markup Language — by hand or with rudimentary tools.

HTML is, in essence, a set of formatting instructions wrapped around your text, like an editor's handwritten marks on a manuscript telling the typesetter "make this a heading," "make this a link." In the 1990s most people writing for the Web actually typed those marks themselves, surrounding their words with angle-bracketed tags. There was no "type into a box and click Post"; the box did not yet exist.

You then had to get those files onto a computer that was permanently connected to the Internet and configured to serve them — a web server.

A web server is, in essence, a shop that is always open: a computer that sits on the Internet around the clock, listening for requests and handing back the requested page. A document on your own desktop computer is not on the Web; it becomes part of the Web only once it lives on a machine that is always reachable at a stable address. The leap from "I wrote a file" to "the world can read my file" was the leap of acquiring, or renting, an always-open shop.

To get files onto that server you generally used FTP (File Transfer Protocol), a separate tool with its own conventions. You needed to understand file paths, permissions, and the relationship between a human-readable web address and the underlying machine. None of this was insurmountable — millions did learn it — but each step was a small barrier, and the barriers compounded. The cumulative effect was that publishing selected for the technically inclined, the institutionally supported, and the determined hobbyist.

You needed infrastructure, and infrastructure cost money

Running a web server meant either owning a suitable always-on machine with a permanent Internet connection — rare and expensive for an individual in the 1990s, an era when most home users reached the Internet through intermittent dial-up connections that tied up the telephone line — or renting space on someone else's server through a web hosting provider, which charged a monthly fee and still required technical configuration. Free services such as GeoCities and Angelfire lowered this barrier dramatically, which is exactly why they attracted millions and why they loom so large in the period's memory. But even the free services demanded that the user learn HTML and the mechanics of uploading, and they constrained what could be built.

The net result was a publishing layer that was open in principle but gated in practice. Anyone was permitted to publish — there was no gatekeeper deciding who could have a website — yet the practical prerequisites of skill, tools, and infrastructure meant that the population of publishers remained a small fraction of the population of readers. This distinction between permissionless in principle and accessible in practice will recur throughout the book, because it is precisely the gap that Web 2.0 closed and the gap whose re-opening, in a new form, Web3 critics and proponents argue about.

In plain terms: the early Web was like a public square where anyone was legally free to stand on a soapbox and speak — but you had to build your own soapbox, from scratch, with tools most people did not own. The freedom was real; the friction was also real; and the friction is what kept the square mostly full of listeners.

The open-protocols ethos and early decentralization

If the "read era" had a virtue that later commentators — especially Web3 advocates — would look back on with something like longing, it was this: the early Web was built on open protocols, and was in a meaningful sense decentralized.

An open protocol is, in essence, a shared public language with published grammar rules that anyone may use without asking permission and without paying a toll. HTTP (how browsers and servers talk), HTML (how pages are described), and the Internet's underlying TCP/IP (how data moves) are all open protocols. Because the rules are public and unowned, anyone can write a program that speaks them — and so anyone can build a browser, or a server, or a website, without a corporation's consent.

The architectural consequence was that the Web had no centre and no proprietor. CERN had placed the core software in the public domain in 1993 (CERN, n.d.). To put a site online, you did not register with a company that could refuse you, throttle you, or harvest your data as a condition of entry; you ran software that spoke the open protocols and you joined a network of peers. There were millions of independent web servers, each operated by a different person or organization, each addressable directly. The architecture was, in the language this book will develop, permissionless: participation required conformance to public rules, not approval from an owner.

Electronic mail exemplified the same property and is worth dwelling on because it remains, to this day, the most successful decentralized communication system ever built. Email is federated.

A federated system is, in essence, like the international postal system or the telephone network: there is no single company that owns "email," yet a message sent from one provider arrives reliably at another. Anyone can run their own mail server and still exchange messages with everyone else, because all the servers agree to speak the same open protocol. Federation is the middle path between one central owner and total fragmentation — many independent operators, one shared language.

A person could run their own mail server, or use a university's, or an Internet service provider's, and still correspond freely with anyone on any other server anywhere in the world. No central authority sat in the middle, owned the social graph, or could unilaterally cut a user off from the whole network. This federated, open-protocol character is the historical reality that the word "decentralization" — which becomes the ideological heart of Web3 (see Chapter 26, Web3 as Ideology: Decentralization, Cypherpunk Roots, and the Politics of Trust) — gestures back toward. Part of the rhetorical force of Web3 comes from the claim that the Web began decentralized, drifted into corporate centralization during the Web 2.0 era, and ought to be returned to its open roots. The early Web's genuine openness gives that narrative real historical purchase, even as later chapters will complicate it.

Two honest qualifications are required so that this analogy does not curdle into myth. First, the early Web was decentralized in its architecture — anyone could run a node — but its traffic and attention were already concentrating, as the rise of portals and the fierce competition to be the user's default starting page demonstrate. Decentralized infrastructure does not guarantee decentralized usage, a lesson that will haunt every era covered in this book, including Web3 itself. Second, decentralization came at the cost of difficulty: the same absence of a central provider that made the Web open is part of why publishing was hard. There was no helpful platform to do the work for you precisely because there was no platform. The convenience that Web 2.0 would later offer, and the centralization it would entail, are two faces of the same coin — a tension this book returns to repeatedly.

The dot-com boom and bust as context

The "read era" did not unfold in a calm laboratory; it coincided with one of the largest speculative financial episodes in modern history, and that economic backdrop is part of what the period was. As the public Web grew through the second half of the 1990s, investor enthusiasm for Internet-related companies — the so-called dot-coms, named for the ".com" web address suffix — inflated into a classic speculative bubble.

In plain terms: a speculative bubble is what happens when the price of something rises far above any reasonable estimate of its underlying worth, driven by the expectation that prices will keep rising and that a greater buyer will appear. People stop asking "what is this business actually worth?" and start asking "can I sell it to someone else for more next month?" The dot-com bubble applied this logic to Internet startups, many of which had enormous valuations and no profits.

Capital poured into companies whose business plans amounted to little more than a website and a promise of future dominance. The technology-heavy stock indices climbed steeply, then collapsed in 2000–2001, wiping out vast sums and bankrupting a large number of firms whose revenues had never justified their valuations. This book does not treat the financial mechanics of the crash in detail, but two points about it bear directly on the larger argument.

First, the boom both funded and distorted the early Web. It accelerated the build-out of infrastructure and brought the Web into mainstream consciousness, but it also produced a flood of thinly conceived ventures and a culture of hype in which the gap between rhetoric and reality grew dangerously wide. Second — and this is the load-bearing connection — the survivors of the crash are what gave "Web 2.0" its meaning. When the term was crystallized at the 2004 O'Reilly conference, the brainstorming that produced it was explicitly an attempt to characterize the companies that had survived the dot-com bust and the qualities that distinguished them (Wikipedia: Web 2.0, n.d.; Wikipedia: Dale Dougherty, n.d.). The periodization of the Web, in other words, is partly a periodization of a financial cycle: "Web 1.0" is, among other things, the Web of the bubble, and "Web 2.0" is the Web of the companies that lived through its bursting.

This pattern — a wave of technological enthusiasm, a speculative financial bubble, a crash, and a subsequent winnowing that reveals which projects had real substance — is one the reader should hold in mind, because the book will argue that the cryptocurrency and Web3 markets have repeatedly exhibited a structurally similar dynamic (see Chapter 29, Critiques and Skepticism: Hype, Scams, and Recentralization in Practice). Recognising the dot-com episode as a template helps the reader assess later claims with neither naïve enthusiasm nor reflexive dismissal: bubbles can be real and leave behind genuinely transformative infrastructure. Both things were true of the dot-com era.

The limitations that motivated the next era

We can now state plainly the constraints of the read era — not as failings to be mocked from the comfort of hindsight, but as the specific frictions whose removal defines the era that followed.

The first limitation was the publishing barrier already described: contributing required technical skill and infrastructure, so the Web remained, for most people, a place to consume rather than create.

The second was statelessness and the absence of easy interactivity. Pages were largely fixed documents; building a site that remembered who you were, accepted your input, and responded to it required custom server-side programming that was beyond most publishers. The rich, responsive, application-like experiences that users would come to expect were difficult and rare.

The third was the discovery problem. With finding aids split between hand-curated directories and immature search engines, locating relevant content among a rapidly multiplying number of sites was genuinely hard. The eventual dominance of high-quality automated search — and the algorithmic organization of information it enabled — was both a solution to this problem and, as later chapters will discuss, a powerful new point of centralization.

The fourth was the community gap. Because the publishing layer was so asymmetric, the early Web was comparatively poor at supporting the dense, many-to-many social interaction that humans crave. Conversation happened in the margins — guestbooks, forums, newsgroups — rather than being woven into the fabric of the pages themselves.

Each of these limitations is, read in reverse, a description of what Web 2.0 would offer: tools that let anyone publish by typing into a box and clicking a button; dynamic, interactive, application-like sites; powerful search; and platforms designed from the ground up around user participation and social connection (O'Reilly, 2005). The next chapter takes up that transformation in detail (see Chapter 5, Web 2.0: The Read–Write Era and the Platform Economy). The essential point here is that Web 2.0 solved real problems. Its conveniences were not frivolous; they brought the Web's latent promise of universal participation within reach of ordinary people for the first time. This matters for intellectual honesty, because the Web3 critique of Web 2.0 (Chapter 6, The Discontents of Web 2.0) is most persuasive when it acknowledges that the centralization it laments was the price paid for genuine, widely valued gains — not a swindle imposed on an unwilling public, but a bargain whose hidden costs only became clear later.

Establishing the baseline: "read," and what it sets up

The deepest reason this chapter exists is that "Web 1.0" functions as the anchor point for the schema through which the Web3 movement narrates the history of the Web and argues for its own necessity. That schema, stated compactly, runs:

The table is a heuristic, not a law, and the book will interrogate every cell of it. But its first column — "Read" — is the one this chapter has filled in, and the integrity of the whole edifice depends on getting it right. Two features of the read era do real work in the later argument.

First, the asymmetry of participation (many read, few write) is the problem Web 2.0 claims to solve, and the slogan "Read–Write" is intelligible only against the backdrop of a "Read" era in which writing was the privilege of the technically capable few. Second, the open, federated, decentralized architecture of Web 1.0 is the lost virtue that Web3 claims to restore. The Web3 story is, in essence, a fall-and-redemption narrative: an open and decentralized Web (1.0) gave way, in exchange for participation and convenience, to a centralized and extractive Web (2.0), and Web3 proposes to recover the original openness while keeping the participation — adding, on top of "read" and "write," the new capacity to "own" (see Chapter 25).

The critical reader should notice immediately that this narrative leans on the very periodization whose retrospective, constructed nature this chapter opened by flagging. The "read / read–write / read–write–own" schema is genuinely useful as a way of organizing a complex history, and this book will employ it — but it is an interpretation offered, in significant part, by parties with a stake in the conclusion. The early Web was not designed as "version 1.0" of anything, and the neat ascent from reading to writing to owning is a story told after the fact, by people for whom "owning" is the punchline. To hold the history and the interpretation apart — to use the schema without being captured by it — is exactly the discipline this book aims to model.

Synthesis

Web 1.0 names, retrospectively, the first roughly decade-and-a-half of the public World Wide Web: a medium of largely static, hand-built pages, organized by human-curated directories and accessed through competing portals, built atop open and unowned protocols, and federated rather than centrally controlled. Its defining social fact was an asymmetry between the many who read and the few who published, an asymmetry produced not by any gatekeeper but by the real friction of needing technical skill, tools, and always-on infrastructure to put a page on the Web. That friction coexisted with a genuine architectural openness — anyone could, in principle, run a server or a mail node and join the network as a peer — and that openness, set against the speculative drama of the dot-com boom and its 2000–2001 collapse, frames everything that follows.

For the argument of this book, the read era supplies the baseline. It establishes the "Read" pole of the read–write–own schema through which Web3 narrates its own justification, and it furnishes, in its open and federated design, the historical memory of decentralization that the Web3 project invokes as both precedent and goal. Recall this book's working definition: Web3 is best understood not as a single technology but as a contested socio-technical project — a bundle of blockchain-based technologies wedded to an ideology of decentralization and digital ownership, proposing to re-architect the web so that users, rather than platforms, control identity, data, and value. The phrase "rather than platforms" only acquires its full force once we have seen a Web that existed before platforms — open, decentralized, and difficult — and once we understand why the platforms arose. That is the subject of the next chapter, in which the Web learns to let everyone write, and, in the bargain, begins to centralize.

References cited in this chapter

CERN. (n.d.). A short history of the Web. CERN. https://home.cern/science/computing/birth-web/short-history-web

DiNucci, D. (1999, April). Fragmented Future. Print, 53(4), 221–222. http://darcyd.com/fragmented_future.pdf

O'Reilly, T. (2005, September 30). What Is Web 2.0: Design Patterns and Business Models for the Next Generation of Software. O'Reilly Media. https://www.oreilly.com/pub/a/web2/archive/what-is-web-20.html

Techopedia. (n.d.). What Is Web 1.0? Definition, History & How It Works. Techopedia. https://www.techopedia.com/definition/27960/web-10

Wikipedia. (n.d.). Dale Dougherty. https://en.wikipedia.org/wiki/Dale_Dougherty

Wikipedia. (n.d.). GeoCities. https://en.wikipedia.org/wiki/GeoCities

Wikipedia. (n.d.). Web 2.0. https://en.wikipedia.org/wiki/Web_2.0