Sep 7, 2006
Computing at CERN in the 1990s
In the first three issues of CNL this year we looked at the history of computing at CERN up to the 1980s. In the nineties the most important event in computing was probably the birth of the World Wide Web (WWW). This issue will focus on articles related to this revolution. The first Web version of CNL appeared in October–December 1993, and you can read it at http://cern.ch/cnlart/214.
We have found Tim Berners-Lee's first articles for CNL. This shows how important it is to write down information that can be archived and retrieved later. We hope that this will encourage people to submit articles to CNL, especially when working on new projects, to describe and document the various phases.
For a detailed explanation of how the WWW project started and why, check out Tim Berners-Lee's article 'HyperText and CERN', dated March 1989 and May 1990 (www.w3.org/Administration/HTandCERN.txt).
A project for hypertext and browsable documents at CERN
"Depending on its complexity and distribution, a document will be presented to the users in one of the following forms:
- on paper (major manuals, or document widely distributed outside CERN);
- as plain text for simple terminals (we hope all documents);
- as a sophisticated version for workstations – these versions will be developed in the context of the hypertext project (currently being launched between ECP and CN).
The major long-term aim will be to make machine readable versions browsable from outside CERN." (July–October 1990)
World Wide Web: online information for everyone
"A world of information is now available online from any computer platform... We summarize the information currently sourced at CERN, and we introduce the World Wide Web (W3) program, which allows you to browse and search all the data in a simple and consistent manner. Details of all the information, and of how to acquire and install the W3 software on your own machine, are available by typing telnet info.cern.ch…
Information… is usually stored on one platform (the machine, the format and the application that controls the information). The platform has peculiarities which make it different from other platforms. However most machines are connected by networks, and most systems have (terminal) emulators through which they give access of some level to other platforms. Therefore, there are three ways to access information:
- go to the platform the data is on, log into it, use the application specific for these data;
- connect to the platform, from another platform, and use it from there; or
- use a platform-independent application.
In the third case, you can navigate through the information without having to know details depending on a platform. In the following section we will show how to invoke the platform-independent application: W3. To date, only W3 can give you platform-independent access…
The browser described here is the simplest interface to the World Wide Web, which was designed to run on any dumb terminal. A hypertext browser/editor is available under NeXTStep. More powerful interfaces are under development, including Macintosh, X-Windows, VM full-screen and emacs…
T Berners-Lee, R Cailliau, J-F Groff, B Pollermann." (October–December 1991)
W3: spring releases
"The World Wide Web bounds into spring with some exciting software releases from outside CERN. A few more servers pop up, and we look at the Gopher and WAIS part of the Web...
Statistics we take on the access to our server show an exponential increase, with a time doubling about every two months, and currently running at around 500 document fetches a day… many of these are from people telnetting to info.cern.ch…
X11 browsers: Erwise For those who have waited and asked so many times, here are not one but two client products which run under X windows… The Erwise browser was developed at Helsinki Technical University… It is a multi-window, multi-font hypertext browser which uses the W3 common access library and so has full W3 functionality…
X11 browsers: Viola The Viola browser by Pei Wei of Berkeley is another interesting product… Pei's solution to the multiwindows/single window dilemma is to do everything in one window by default, with a 'clone' button which duplicates the current window…
W3 people at CERN Jean-François Groff, who has been a major mover in the W3 architecture, and has been invaluable in keeping the show on the road, came to an end of his contract at the end of March." (March–April 1992)
New version of Mosaic available
"Mosaic is an X11-based application to browse information stored in the World Wide Web system. It has been developed at NCSA and provides at present the best user interface available on Unix workstations.
We have released into the Public Area the Unix, VMS and Macintosh versions of Mosaic 2.2.
In addition Mosaic is available on PC/Novell as provided by Robert Cailliau/ECP…
PCWeb is in fact the SUN version of Mosaic, displaying on PCs which have access to the XVision product. There is an icon called World Wide Web in the Comms folder inside Program Manager. The inconvenience of this version… is that due to the way it has been set up, Mosaic will NOT remember the documents you have 'visited'." (January–March 1994)
Using CERN's phone book from W3/Mosaic
"Two kinds of documents in the CERN-wide Web are of interest…
1. documents informing people about software, projects services etc.; and
2. documents informing people about people, especially their e-mail addresses and phone numbers.
In the first case a reference to the CERN phone book is needed for the signature which indicates the author or person responsible for the document or service… In both cases it is useful to have a direct pointer to a particular person… This identifier, called ccid ( Computer Centre IDentifier), was introduced some years ago and can be used now also in W3/Mosaic." (April–June 1994)