Development and Evolution of HTML

You might try a fun little experiment to see how pervasive Web pages have become. Pick any dictionary. Pick any word, even at random. Type it in your search engine. You will very, very, very likely find a Web site devoted to that word. The Internet is not synonymous with the World Wide Web, but the Internet drove it, and then it drove the Internet as use synergistically increased the demand for network bandwidth, servers, software technologies, and hardware. The mantra for the 1990s would be ‘‘bigger, better, cheaper, faster.’’ And the gurus of the Web and the Internet delivered, every year: better, faster, cheaper, but not, in vindication of Moore’s Law, necessarily bigger.

Gordon Moore, an engineer and one of the founders of Intel (along with Robert Noyce), observed that each year the number of transistors on a board doubled at the same price every 18 months. That observation has since become known as Moore’s Law, and it has been applied to the idea that computer CPU capacity doubles every 18 months, while the price remains steady. One of the earliest lessons of the Internet Age was that software drives hardware development, or, as one professor more delectably phrased it, ‘‘Software makes hardware happen.’’

Hypercard

Hypertext (that is, text that can be used as links to other data, or that responds in some way to a mouse click or hover such as having a caption appear above an icon explaining its function had been around for a while in theory and in practice. Apple came out with Hypercard in 1986. Hypercard was based on earlier concepts out of the 1940s, which had ‘‘cards’’ that held fields holding unique data. These cards were almost like early iterations of forms that could be used to store and manipulate data.

These cards were linked to other cards and kept in a stack that could be accessed by a previous/ next feature. Think of your standard 3 × 5 index card or Rolodex card. The cards all look the same, but the information each card contains is unique. You can put these cards in any order. You can toss them if you want, and write data on a new card, or you can just cross out old data and write new data on the old card. Suppose you have a CD game collection and you decide to keep information on each of your CDs on index cards that you keep in a small box on your desk (next to your desktop). The cards all have the same format: Name of CD, Group, Game type, Year, and Company. Each of those items of information is located on a separate line. The values input to those fields (CD, Group, Game type, Year, Company) would be unique (unless your had two or more copies of the same CD), but the format of the cards would be identical. This foreshadowed the format/content dichotomy central to HTML. The technology of Hypercards, or hypertext, was naturally applied to the links that were developed within HTML coding.

HTML

The coding that drives the Web is HTML, and even though we are moving toward more general implementation of HTML version 4.01 and XHTML, the most browser-safe version is still HTML 2.0, because almost all graphical browsers support it. HTML 4.01 includes enhanced scripting support, stylesheet support, better multimedia support for streaming elements, and additional control over layout of the page.

Even though the latest HTML is version 4, it did not really have four generations; it actually had just three (including this latest evolution. 4.01 and 4.02). The first standardized (emphasis on standardized) version was HTML 2.0. It incorporated support for interactive form elements, including text boxes, labels, Option and Submit buttons, clear and exit, and table structure.

Leading up to 2.0 was a series of rapid developments regarding design features for Web implementation scattered around so much that, until 1993, no one really thought to put together. While the father of HTML (and HTTP and the browser) was Berners-Lee, his creation was rooted in some fertile soil.

His innovations were twofold:

- The concept of linking documents so that a user could look at documents from several different sources on his or her computer

- Modeling HTML on an existing worldwide standardized markup language Standardized General Markup Language (SGML).

From SGML, Berners-Lee borrowed several of the tags we will use in article 4 to make a Web page. From SGML also came the concepts of a document, of separating a document into blocks of data, and of tagging (or marking up) the blocks of data and the elements contained therein. The genius of Berners-Lee was to visualize and then create connections where none had existed before. He took hypertext and married it to a linking technology to give people the ability to bring blocks of data from various referenced sources on the screen. They had never been able to do this before. So, in a manner of speaking, Berners-Lee could be seen as the link that made HTML the phenomenally successful Web design language it is today. In other words, succinctly put, he invented the link.

It might be argued, however, that David Raggett of Hewlett-Packard was a cofounder, for he so much improved upon this basic structure after a visit with Berners-Lee in 1992. Though we won’t go into it the story of his lonely efforts at home night after night to create what would amount to a new language (HTML+) on a crowded dining room table, his dogged pursuit to present a language that addressed Web developers’ common needs should be inspiring to all who stubbornly labor long hours in obscurity to bring to completion the fruit of a great idea. So, by 1992 we have the second iteration (or first codification) of HTML, HTML+, which contained the following features:

- The capability to flow text around images and forms with captions

- The resizing of tables

- The creation of image backgrounds

- The use of arithmetical and logical relational operators

Shows an example of these features. Raggert demonstrated all of these features in 1993 at the first http://www Conference by using the Arena browser he had developed to showcase HTML+. One of the significant results of the http://www Conference was the decision to have one platformindependent version of HTML that could be used by any browser. At the conference, it was agreed that the work on HTML+should be carried forward to lead to the development of an HTML 3 standard. Unfortunately, the rush of time and technology would not allow standardization to take effect, and the growth and acceptance of later versions of HTML was somewhat hindered by the competing browsers used to render Web pages. This is why you have compatibility issues and must test for appearance of your Web page in multiple browsers. The big two, until quite recently, were Netscape versus IE. IE appears to have won the majority of market share. However, a successor to the Netscape browser, FireFox, is quickly gaining market share, and Apple’s Safari browser is the default for the resurgent Macintosh market. The point is this: Test on the most popular browsers before you deploy your Web site, or users will complain.

Some of Raggett’s planned features were scooped by the talented team from NCSA Mosaic led by Eric Bina. They were the first to present added images, nested lists, and interactive forms. And sometimes being first is enough, though some in academia and development circles have never ceased complaining that these features eventually incorporated into the corpus of HTML by version 3 weren’t all that great.

Because of the lack of cohesive development structure (that is, a standards body presenting a template of development for Web authors to follow), the support for HTML has been historically uneven. We have seen the addition of features (generically called extensions) that competing browsers put out that would work only on their own browsers. The use of Layers in Netscape is one illustration, and the coding differences in 3D shading for borders is another.

Cascading Style Sheets (CSS), used to simplify and standardize the formatting of Web sites and integrated into HTML 4.01, actually got its start as an extension. It would take years for popular code changes to percolate through browser versions and be widely accepted. Others would gradually pick up popular scripts, say, useful to sending data to a printer, and then someone else would write a piece of code that was useful and it would also filter into the user community. By the mid-1990s, you had an amorphous mass of programs that would flow in and out of a programmer’s toolkit, and no real standardized set of tools that everyone could use all the time to accomplish set tasks.

The 1996 effort at standardization, HTML 3.0, just never really took off. Yet, in 1997, a newer implementation of the standards set forth in HTML 3.0 was again offered. HTML 3.2 offered increased support for tables, and gave more elements to interactive forms. But what may have underlain its importance (and acceptance) was the definition of the language allowing the incorporation of scripts as a Web authoring tool, not to mention backward compatibility.