I'm making a small documentary about the invention of the World Wide Web and am reading Tim Berners-Lee's book Weaving the Web. I'm confused by a passage describing how his browser could follow links to Internet news articles and newsgroups, even though they were not hosted using HTTP.
My understanding is that Usenet articles were distributed between servers using NNTP and stored locally in news spools. The passage says that Berners-Lee's browser made these articles available as hypertext pages and claims they were transferred using FTP. From a user's point of view, how did this actually work? Did the browser retrieve articles through NNTP, access FTP-based archives, or use some kind of gateway or translation layer? Did the articles themselves become hypertext, or did the browser simply display existing Usenet content in its document interface?
3 Answers
Think of it as adding a universal front end rather than rewriting every source. A news link could identify an article or newsgroup, and the browser would use the appropriate mechanism to fetch it, then render the result for the user. The significance was that Web navigation could sit on top of information that already existed elsewhere on the Internet.
The important idea is that the browser was designed to handle more than just HTTP URLs. It could use different protocols or resource handlers, so a user could click a link to a news article and have the browser retrieve and display it alongside ordinary Web pages. That did not convert Usenet into HTTP or change the article itself into a Web document; it made an existing Internet information system accessible through the same browser interface.
The FTP reference is probably about material that had been made available in FTP archives, or about a particular way the early browser accessed news-related content. FTP predates HTTP and was commonly used to retrieve files, while NNTP was the protocol associated with Usenet news servers. Early Internet resources were often mirrored or gatewayed between systems, so the browser could present content from several protocols as navigable documents.
In other words, the browser supplied the common presentation and navigation layer. The underlying server and protocol could still be different depending on the link.
Usenet distribution was generally handled by NNTP, so FTP should not be read as the normal protocol used to exchange live Usenet articles.