TML, that universal and egalitarian documentation format, is now 13 years removed from birth and 10 years removed from the lab. Unless pretenders such as Adobe's PDF quicken their pace, HTML documents will remain the dominant format for publicly accessible content for some time. Few question the egalitarian value of HTML's center-stage position, but for technologists the spotlight is clearly wandering elsewhere.

A Victim of Its Own Success
HTML has been successful—too successful. That's a well known problem typical of mature systems, regardless of the simplicity or naïveté of their origin. That success has led to a growing number of layers built on top of, and dependent upon HTML (see Figure 1).

Figure 1: Crushing HTML. Success has left HTML at the bottom of a growing number of layers built on top of, and dependent upon it.

When HTML was new, it was the king of Web document formats, and it grew and changed rapidly to accommodate increasingly complex documents. Unfortunately, the overwhelming weight of the layers built on top of HTML has forever fixed it in its current, standardized, unchangeable form. It is a target document format for many power tools and those power tools are not likely to change. It is a source document format for browsers, and browsers will support it in it current form forever. And any future it might have had is now occupied by XML.

Having standards experience is useful and marketable. Having standards experience that is being flanked by other, newer standards is less so. Regardless of its usefulness, HTML's inability to change is a warning signal: the game has moved on. Anyone tied to HTML should take notice.

Unfortunately, HTML has two characteristics that make it extremely difficult to change: it's simple, and it's visible. Because people can see the markup tags in a text editor or on a screen, and read them as pseudo-English words, they're considered important. What they don't see is that the markup scheme itself is dispensable—it's the results that they should care about. Increasingly, producing HTML has changed from manually editing text to favor WYSIWYG editors, where the HTML creation process is hidden.

That is not to say that developers should now ignore HTML altogether. Good C programmers are concerned with the quality of the assembly code generated from their source. Good Web developers should be equally concerned with the quality of the HTML generated from their tools. Nevertheless, the need to assimilate the finer points of the HTML standard isn't the priority it once was; HTML has been subsumed by other tasks. Those finer points are now the province of a limited few who are paid to maintain HTML standards compliance.