Archie was the first search engine, built in 1990 to index files on computers connected to the early internet

Archie launched in 1990 as the first tool that let you search for files across multiple computers on the internet. It was created by Alan Emtage, a student at McGill University in Montreal, who built it to solve a real problem: the internet had thousands of computers sharing files, but no way to find them. Archie worked by automatically visiting computers, cataloging what files they had, and letting you search that catalog from your own machine.

The name came from "archive" — Archie was literally an archive you could search. It did not search the content of web pages (the web barely existed yet). Instead, it found files by their names and the computers that held them. If you wanted a specific piece of software or a document, you typed the filename into Archie, and it told you which computers had it and how to read it.

Archie worked because the internet in 1990 was small enough to index. There were maybe 3,000 computers connected. Archie's server would connect to each one, ask what files it was sharing, write down the names, and move to the next. When you searched, you were searching that stored list, not the live internet. This meant Archie's index was always a few days or weeks out of date — but it was still infinitely better than having no way to search at all.

Key Takeaways

  • Archie, built in 1990 by a McGill University student, was the first search engine and indexed filenames on computers across the early internet.
  • Archie searched for files by name only, not by the content inside them, because the web and web pages did not yet exist.
  • The tool worked by automatically visiting thousands of computers, recording what files they shared, and letting users search that stored list.
  • Archie solved a real problem: the early internet had no way to find files, even though thousands of computers were sharing them.
  • Later search engines like Veronica, Jughead, and eventually web-based engines like WebCrawler and AltaVista replaced Archie as the internet grew.

How Archie actually worked: the mechanics of the first search

Archie ran on a single server at McGill University. Every night, it would connect to computers that were publicly sharing files — usually university computers and research institutions. It would ask each one: "What files do you have?" The computer would send back a list of filenames, and Archie stored all of those names in a database on its own server.

When you wanted to search, you connected to Archie using a straightforward text interface (there was no web browser yet — you used a program called a Telnet client to connect). You typed a filename or part of a filename. Archie searched its database and showed you a list of computers that had a file matching that name, along with the exact path where the file lived. You then used a separate tool called FTP (File Transfer Protocol) to read the file from that computer.

This was clunky by today's standards, but revolutionary at the time. Before Archie, if you wanted a file, you had to know which computer had it — or you had to ask someone who knew. Archie made it possible to search across the entire connected internet from one place.

Why Archie worked in 1990 but would never work today

Archie could only work because the internet was tiny. In 1990, there were roughly 3,000 computers connected to the internet worldwide. Archie could visit all of them in a night or two and have a complete index. Today there are billions of devices connected, and the amount of data is incomprehensibly larger. A modern search engine like Google does not visit every computer — it crawls web pages, which are designed to be indexed, and it uses algorithms to rank the most relevant results.

Archie also only searched filenames, not content. It could tell you that a file called "budget.xls" existed on a computer in Boston, but it could not tell you what was inside that spreadsheet. This worked fine when most internet users were researchers and students looking for specific software or datasets. But when the web became visual and the internet became public, people wanted to search for ideas and information, not just filenames.

The web itself — invented by Tim Berners-Lee in 1989 but not widely used until the mid-1990s — made Archie obsolete. Web pages were designed to be linked and indexed. Search engines like WebCrawler (1994) and AltaVista (1995) could read the actual content of pages and let you search by topic or keyword. By the late 1990s, Archie was gone.

What came after Archie: Veronica, Jughead, and the web search engines

Archie's success inspired other tools. Veronica (1992) and Jughead (1993) were search engines for Gopher, a different system for sharing files that competed with the web in the early 1990s. Both worked on the same principle as Archie — they indexed filenames and let you search them — but they were built for Gopher's structure instead of the raw internet.

The real replacement came when the web took over. WebCrawler, launched in 1994, was the first search engine to index the actual content of web pages. It would follow links from page to page, read the text on each one, and let you search by keyword. AltaVista (1995) did the same thing but faster and with better results. Yahoo (1994) started as a hand-organized directory of websites but added search. By the late 1990s, Google (1998) arrived with a ranking system that made search results far more useful than its competitors.

Each of these tools solved the same problem Archie solved — how do you find something on a network of millions of computers? — but they worked on a vastly larger scale and searched content instead of just filenames.

Why Archie matters even though nobody uses it anymore

Archie proved that automated search was possible and useful. Before Archie, most people assumed you would navigate the internet by asking someone or by browsing directories that humans had organized by hand. Archie showed that a computer could automatically visit other computers, collect information, and let you search it. That idea became the foundation for every search engine that came after.

Archie also showed the limits of what you can index. The internet grew too fast for Archie to keep up. This taught later search engines that they needed to be smarter — they could not visit every computer, so they had to be selective about what they crawled, and they had to rank results by relevance instead of just listing everything that matched. Google's entire business model came from solving the problem that killed Archie: how do you search a network so large that you cannot index all of it?

The people who built Archie and why they built it

Alan Emtage, a computer science student at McGill University, built Archie in 1990 as a solution to a frustration he and his classmates shared. They knew files existed on computers across the internet, but there was no way to find them. Emtage wrote a program that would automatically log into computers, read their file listings, and store them in a searchable database. He called it Archie and made it available to other internet users.

Emtage did not try to commercialize Archie or patent it. He released it as a tool for the research community. This was typical of early internet culture — people built tools to solve problems and shared them freely. Emtage went on to work in internet infrastructure and security, but Archie remains his most famous contribution.

The fact that Archie was built by a student, not a company, is important. The early internet was built by researchers and students solving problems they actually had. There was no business model, no venture capital, no plan to make money. Archie existed because Emtage needed it and had the skills to build it.

How search engines work today compared to Archie

Modern search engines like Google work on principles that Archie invented but at a scale Archie could never reach. Google crawls billions of web pages, reads their content, and indexes them. When you search, Google does not search the live web — it searches its index, just like Archie did. But Google's index is updated constantly, covers far more content, and uses algorithms to rank results by relevance.

Google also does not visit every computer on the internet. Instead, it follows links from page to page, starting from known websites and following the connections. This means some pages get indexed and some do not, depending on whether they are linked to from other pages. Archie had the opposite problem: it tried to index everything but could only reach computers that were explicitly sharing files.

The core idea is the same: a computer automatically visits other computers, collects information, stores it in a database, and lets you search that database. Archie proved this idea worked. Everything that came after — WebCrawler, AltaVista, Google, Bing — is a refinement of what Archie started.

Frequently Asked Questions

Did people actually use Archie, or was it just a prototype?

Archie was genuinely used by thousands of internet users in the early 1990s. It was the only tool that let you search for files across the internet, so anyone who needed to find software or data used it. Universities and research institutions relied on it. It was not a prototype — it was the actual search engine of the early internet.

Could Archie search the web, or only files?

Archie could only search filenames on computers that were sharing files through FTP. It could not search web pages because the web barely existed when Archie was built. By the time the web became popular in the mid-1990s, Archie was already being replaced by web-based search engines.

Why did Archie fail if it was the first search engine?

Archie did not fail — it became obsolete. It worked perfectly for its purpose, but the internet changed. The web became the dominant way people shared information, and web pages were designed to be indexed by search engines. Archie's approach of visiting every computer and indexing filenames could not scale to billions of web pages. Newer search engines like WebCrawler and Google solved the problem of searching the web instead of files.

Is Archie still running anywhere?

Archie is no longer in active use. The original server at McGill University shut down in the early 2000s. Some historical archives have preserved Archie as a piece of internet history, but it is not used for actual searching anymore. The tools that replaced it — web search engines — are far more useful and powerful.

How did people find things on the internet before Archie?

Before Archie, there was no automated way to search. People asked colleagues, read mailing lists, or consulted hand-organized directories. Some institutions maintained lists of what files they were sharing, but there was no central place to search across multiple computers. Archie was the first tool that solved this problem automatically.