Tim Berners-Lee and the Web
The quotes here are from: Weaving the Web by Tim Berners-Lee © 1999, Harper, ISBN 0-06-251586-1
The Enquire Program
In 1980 Tim Berners-Lee joined the staff of the physics CERN facility in Switzerland as a contract programmer. In his odd moments he began writing a program he called Enquire. Berners-Lee used Enquire to keep track of who had written which program and on which machine the program ran. He writes: "... I could type in a page of information about a person, a device, or a program. Each page was a "node" in the program ... The only way to create a new node was to make a link from an old node ..."
For each link he created with Enquire he also described the links with information such as who wrote some program or used it. This program only ran on one computer, the central development computer. It was not on a network. The internet which existed would not be used at CERN for years yet. When Berners-Lee left CERN after his contract job was up he gave the disk to engineers but it was lost.
Back to CERN
When Tim Berners-Lee returned to CERN in 1984 he realized that there was a need for a good program in the documentation-systems category. At this time the world of computers was still extremely varied without many compatibilities. Even computers running MS-DOS ran non-compatible versions of MS-DOS. But he saw this as on opportunity. He would need to construct a program which would combine the links of Enquire with hypertext and it should be able to link to anything anywhere. The program should be de-centralized, no central database to hold the links and no admin for access permissions. This way the program could scale to any size without getting bogged down in overhead. By late 1988 he put in a proposal to get a hypertext system going at CERN.
Ben Segall, a former mentor had worked in the US and was introduced to the internet there. He was an advocate for using the internet at CERN. Before the internet and before networks computers were hooked to terminals or other computers with individual wires. This greatly limited the number of computers which could communicate with each other. The solution was a network. Networks indirectly link computers by sending information across the same with but addressed to specific computers. It is the difference between hiring a postal carrier for each house or hiring one postal carrier to drop letters at their proper addresses on that postal route. On networks a protocol handles these details.
The original NeXT machine which ran the first ever web server. The World Wide Web was developed on this machine. Copyright CERN - Click for more pictures from CERN and the birth of the web.
By May 1990 his proposal was still not happening at CERN. He re-did the proposal and it was shelved again. That is when the NeXT hit the market (1989, originally sold for $6,500). He persuaded CERN to buy a NeXT and got the chance to write his hypertext program idea using the NeXT. This would be under the category of checking out the NeXT operating system and the programming capablities.
The World Wide Web
Once he started he began the name of each program with "HT" (for hypertext). Then Berners-Lee decided on a name for his system. He writes: "I would call my system the 'World Wide Web.'" That September he attened the European Conference on Hypertext Technology.
You would think this would be the most receptive audience but the existing hypertext community had a hard time imagining hypertext used universally across the globe. Their hypertext had links within databases or within single computers or systems. Some, such as Dynatext, for electronic books, was compiled (to machine language). They couldn't imagine sending mere text across the globe without going through a compile phase. Worse they insisted on having a central-link database. This would forever limit the number of links possible.
Berners-Lee: "... Their vision was limited to sending text that was fixed and consistent -- in this case, whole books. I was looking at a living world of hyypertext, in which all the pages would be constantly changing. It was a huge philosophical gap."
A Browser, a Server and HTML
In October 1990 he began writing the code for the web client (browser) on the new NeXT computer. The browser was both a browser to display the text and an editor to create the text. He had this in mid-November. He also wrote HTML - Hypertext Markup Language - derived from the more involved SGML (standard generalized markup language). At the same time he wrote the first web server.
The browser could follow HTTP links but he also did something very useful. He added the ability to access huge amounts of information already on the internet from news groups and in articles by allowing his browser to follow FTP (file transfer protocol) links. The news groups were on FTP server.
Robert Caillau - Copyright CERN - Click for more pictures from CERN and the birth of the web.
By Christmas day 1990 it was up and running as a system. Two people on the WWW, Tim Berners-Lee and Robert Caillau, a fellow worker at CERN and a ready collaborator. Caillau made things happen with people connections and resources. Now they needed a reason to start promoting it at CERN.
The Phone Book
They persuaded Bernd Pollermann who maintained the CERN phone book database on an aging mainframe at CERN. Each computer needed a specialized program to access the phone book. But once they could put a web server on the mainframe and then get people around CERN to add the new browser to their machines, anyone could get to the listings in the phone book without special coding.
It worked for everyone but the concept of global wasn't there yet. People still wondered why you couldn't just write a simple program for each machine to download the phone directory. It was the old way of doing things but the dependence on specific software for each machine was very limiting and required multiple versions of any program. Still, the phone directory was a foot in the door and a demonstration of a common framework for any machine anywhere in the world.
In March of 1991 Berners-Lee released his WorldWideWeb program to CERN personnel with NeXT computers. From there a visiting Stanford physicist, Paul Kunz arrived with his NeXT and on his return to Palo Alto introduced the Web to Louise Addis, the librarian at the Stanford Linear Accelerator. She saw the web as a godsend solution to get out of their sophisiticated but mainframe-bound system. This would make SLAC's documents available to physicists anywhere.
In August 1991 he put the software up for grabs on his web server at CERN. In July and August of 1991 there were between ten and 100 pages viewed per day. In December 1991 Caillau and Berners-Lee attended the hypertext community meeting in San Antonio. There they managed to cobble together a modem connection through the hotel to the University of Texas to dial in to the computer back at CERN in Switzerland. They were the only persons in the conference who were doing actual connectivity. Everyone else was still limiting their hypertext to central databases and local files. Two years later, in 1993, notes Berners-Lee, "every project on display would have something to do with the Web."