History

The History of HTML: Evolution and Versions

By the CodeCompiler Team · 10 min read

HTML is so deeply embedded in how we think about the web that it's easy to forget it was invented by a specific person, at a specific moment, to solve a specific problem. Understanding where HTML came from — and how it changed over more than three decades — makes it much easier to understand why the language looks the way it does today, including some of its stranger quirks and legacy behaviors.

1989–1991: A Proposal for Linked Documents

In 1989, Tim Berners-Lee, a researcher at CERN, the European particle physics laboratory, proposed a system for linking and sharing documents between researchers over a network. The goal was practical, not visionary in the way we think of "the internet" today: scientists at CERN worked with huge amounts of documentation, and there was no easy way to cross-reference and link related documents together. Berners-Lee's proposal combined an early version of HTML for describing documents, HTTP for transferring them, and URLs for addressing them — the three technologies that, together, became the World Wide Web.

The very first version of HTML was extremely simple: a small set of tags for headings, paragraphs, and links, borrowing conventions from an existing document markup standard called SGML. It had no concept of images embedded directly in a page, no tables, no forms — just structured, linked text.

1993–1995: HTML Goes Public and Starts to Fragment

As the web spread beyond CERN and academic networks, competing browser makers began adding their own custom tags to HTML, without any shared standard governing what should be added. Netscape Navigator, the dominant browser of the mid-1990s, introduced tags like <center> and <font> that only worked in Netscape. Microsoft's Internet Explorer introduced its own set of proprietary extensions. This period, sometimes remembered as the "browser wars," created serious headaches for developers, who often had to build entirely different versions of the same page depending on which browser a visitor used.

HTML 2.0, published in 1995 by the Internet Engineering Task Force, was the first version of HTML standardized as a formal specification, an attempt to document the common ground that had emerged across competing browsers.

1997–1999: HTML 3.2 and HTML 4.01

The World Wide Web Consortium (W3C), formed in 1994 with Tim Berners-Lee as its director, took over stewardship of HTML from this point forward. HTML 3.2, released in 1997, formally incorporated features like tables and applets that had become common in practice. HTML 4.01, released in 1999, was a major milestone: it introduced strong support for stylesheets (encouraging developers to separate presentation into CSS rather than embedding styling directly in HTML), improved support for scripting, and better accessibility features. HTML 4.01 remained the dominant standard for web pages for nearly a decade.

2000–2006: The XHTML Detour

In the early 2000s, the W3C attempted to steer HTML toward stricter, XML-based syntax rules with XHTML — requiring all tags to be properly closed, all attribute values to be quoted, and documents to be well-formed according to strict XML parsing rules. The idea was appealing in principle: stricter rules should mean more predictable, more interoperable documents. In practice, adoption was mixed, and many real-world "XHTML" pages were served and rendered by browsers in a much more forgiving way than the specification technically required, undermining the strictness that was the whole point.

2004–2014: The Road to HTML5

Frustrated with the direction of XHTML, browser vendors including Mozilla, Apple, and Opera formed an independent group called the WHATWG (Web Hypertext Application Technology Working Group) in 2004, focused on evolving HTML based on actual, practical needs of web developers rather than academic purity. This work eventually became HTML5, formally published as a W3C recommendation in 2014, though browsers had already been implementing pieces of it for years beforehand.

HTML5 was a genuinely significant leap forward. It introduced semantic elements like <article>, <section>, <nav>, and <header>; native support for audio and video without third-party plugins like Flash; the <canvas> element for drawing graphics directly with JavaScript; native form validation attributes; and countless new browser APIs for things like offline storage, geolocation, and drag-and-drop.

2019–Today: HTML as a "Living Standard"

In 2019, the W3C and WHATWG agreed that the WHATWG's continuously updated "Living Standard" would be the single, authoritative version of HTML going forward, rather than maintaining two parallel and occasionally conflicting specifications. This means HTML today isn't versioned the way it used to be — there's no "HTML6" on the horizon. Instead, the specification is updated incrementally and continuously, with new elements and APIs added as they mature and gain browser support, rather than bundled into large, infrequent version releases.