Chapter 18. The CSS Saga

The saga of CSS starts in 1994. One of the authors of this book works at CERN the cradle of the Web and the Web is starting to be used as a platform for electronic publishing. One crucial part of a publishing platform is missing, however: There is no way to style documents. For example, there is no way to describe a newspaper-like layout in a Web page. Having worked on personalized newspaper presentations at the MIT Media Laboratory, Håkon saw the need for a style sheet language for the Web.

Style sheets in browsers were not an entirely new idea. The separation of document structure from the document's layout had been a goal of HTML from its inception in 1990. Tim Berners-Lee wrote his NeXT browser/editor in such a way that he could determine the style with a simple style sheet. However, he didn't publish the syntax for the style sheets, considering it a matter for each browser to decide how to best display pages to its users. In 1992, Pei Wei developed a browser called "Viola," which had its own style sheet language.

However, the browsers that followed offered their users fewer and fewer options to influence the style. In 1993, NCSA Mosaic, the browser that made the Web popular, came out. Stylewise, however, it was a backward step because it only allowed its users to change certain colors and fonts.

Meanwhile, writers of Web pages complained that they didn't have enough influence over how their pages looked. One of the first questions from an author new to the Web was how to change fonts and colors of elements. At that time, HTML did not provide this functionality and rightfully so. This excerpt from a message sent to the www-talk mailing list early in 1994 gives a sense of the tensions between authors and implementors:

In fact, it has been a constant source of delight for me over the past year to get to continually tell hordes (literally) of people who want to -- strap yourselves in, here it comes -- control what their documents look like in ways that would be trivial in TeX, Microsoft Word, and every other common text processing environment: "Sorry, you're screwed."

The message is available from the archives at http://www.webhistory.org/ www.lists/www-talk.1994ql/0648.html

The author of the message was Marc Andreessen, one of the programmers behind NCSA Mosaic. He later became a co-founder of Netscape, a company eager to fulfill the request of authors. On October 13, 1994, Marc Andreessen announced to www-talk that the first beta release of Mozilla (which later turned into Netscape Navigator) was available for testing. Among the new tags the new browser supported was CENTER, and more tags were to follow shortly.

The original is online at http://www.w3.org/People/howcome/p/cascade.html

Three days before Netscape announced the availability of its new browser, Håkon published the first draft of the Cascading HTML Style Sheets proposal. Behind the scenes, Dave Raggett (the main architect of HTML 3.0) had encouraged the release of the draft to go out before the upcoming "Mosaic and the Web" conference in Chicago. Dave had realized that HTML would and should never turn into a page-description language and that a more purpose-built mechanism was needed to satisfy requirements from authors. Although the first version of the document was immature, it provided a useful basis for discussion.

The Argo browser was part of a project to make the Internet accessible to scholars in the Humanities. It featured plug-ins (which it called "applets") before Netscape added them. See http://odur.let.rug.nl/~bert/stylesheets.html and http://www.let.rug.nl/~bert/argo.html.

Among the people who responded to the first draft of CSS was Bert Bos, the co-author of this book. At that time, he was building Argo, a highly customizable browser with style sheets, and he decided to join forces with Håkon. Both of the two proposals look different from present-day CSS, but it is not hard to recognize the original concepts.

One of the features of the Argo style language was that it was general enough to apply to other markup languages in addition to HTML. This also became a design goal in CSS and "HTML" was soon removed from the title of the specification. Argo also had other advanced features that didn't make it into CSSI, in particular, attribute selectors and generated text. Both features had to wait for CSS2.

Robert Raisch's message to www-talk is at http://www.webhistory.org/ www.lists/www-talk.1993q2/0445.html

"Cascading Style Sheets" wasn't the only proposed style language at the time. There was Pei Wei's language from the Viola browser, and around 10 other proposals for style sheet languages were sent to the www-talk and www-html mailing lists. Then, there was DSSSL, a complex style and transformation language under development at ISO for printing SGML documents. DSSSL could conceivably be applied to HTML as well. But, CSS had one feature that distinguished it from all the others: It took into account that on the Web, the style of a document couldn't be designed by either the author or the reader on their own, but that their wishes had to be combined, or "cascaded," in some way; and, in fact, not just the reader's and the author's wishes, but also the capabilities of the display device and the browser.

As planned, the initial CSS proposal was presented at the Web conference in Chicago in November 1994. The presentation at Developer's Day caused much discussion. First, the concept of a balance between author and user preferences was novel. A fictitious screen shot showed a slider with the label "user" on one side and "author" on the other. By adjusting the slider, the user could change the mix of his own preferences and those of the author. Second, CSS was perceived by some as being too simple for the task it was designed for. They argued that to style documents, the power of a full programming language was needed. CSS went in the exact opposite direction by making a point out of being a simple, declarative format.

WWW3, the third conference in the WWW series, was held on April 10 14, 1995, in Darmstadt, Germany

At the next WWW conference in April 1995, CSS was presented again. Both Bert and Håkon were there (in fact, this was the first time we met in person) and this time, we could also show implementations. Bert presented the support for style sheets in Argo, and Håkon showed a version of the Arena browser that had been modified to support CSS. Arena had been written by Dave Raggett as a testbed for new ideas, and one of them was style sheets. What started out as technical presentations ended up in political discussions about the author-reader balance. Representatives from the "author" side argued that the author ultimately had to be in charge of deciding how documents were presented. For example, it was argued that there may be legal requirements on how warning labels had to be printed and the user should not be able to reduce the font size for such warnings. The other side, where the authors of this book belong, argued that the user, whose eyes and ears ultimately have to decode the presentation, should be given the last word when conflicts arise.

To subscribe to the mailing list or search the archives, see http://lists.w3.org/Archives/Public/www-style

Outside of the political battles, the technical work continued. The www-style mailing list was created in May 1995, and the discussions there have often influenced the development of the CSS specifications. Now, almost 10 years later, more than 16,000 messages exist in the archives of the mailing list.

In 1995, the World Wide Web Consortium (W3C) also became operational. Companies were joining the Consortium at a high rate and the organization became established. Workshops on various topics were found to be a successful way for W3C members and staff to meet and discuss future technical development. It was therefore decided that another workshop should be organized, this time with style sheets as the topic. The W3C technical staff working on style sheets (namely the two authors of this book) were now located in Sophia-Antipolis in Southern France where W3C had set up its European site. Southern France is not the worst place to lure workshop participants to, but because many of the potential participants were in the U.S., it was decided to hold the workshop in Paris, which is better served by international flights. The workshop was also an experiment to see if it was possible for W3C to organize events outside the U.S. Indeed, this turned out to be possible, and the workshop was a milestone in ensuring style sheets their rightful place on the Web. Among the participants was Thomas Reardon of Microsoft, who pledged support for CSS in upcoming versions of Internet Explorer.

At the end of 1995, W3C set up the HTML Editorial Review Board (HTML ERB) to ratify future HTML specifications. Because style sheets were within the sphere of interest of the members of the new group, the CSS specification was taken up as a work item with the goal of making it into a W3C Recommendation. Among the members of the HTML ERB was Lou Montulli of Netscape. After Microsoft signalled that it was adding CSS support in its browser, it was also important to get Netscape on board. Otherwise, we could see the Web diverge in different directions with browsers supporting different specifications. The battles within the HTML ERB were long and hard, but CSS level I finally emerged as a W3C Recommendation in December 1996.

In February 1997, CSS got its own working group inside W3C and the new group set out to work on the features which CSSI didn't address. The group was chaired by Chris Lilley, a Scotsman recruited to W3C from the University of Manchester. CSS level 2 became a Recommendation in May 1998. Since then, the group has worked in parallel on CSS3 (which is still to come) and CSS 2.1 (which is described in this book).

The W3C working group, whose official name is "Cascading Style Sheets and Formatting Properties Working Group," because they do more than just CSS, has about 15 members, delegated by the companies and organizations that are members of W3C. They come from all over the world, so the "meetings" are usually over the phone, and last about an hour every week. About four times each year, they meet somewhere in the world. Recent venues have been Provo, Redmond, San Francisco, and Paris. In Paris, the meeting was held at the offices of EDF-GDF, the French electricity and gas company. At that meeting, the group was offered a superb dinner: French cuisine overlooking Paris and the Seine one of the few glamourous moments in the history of a hard-working technical working group.