HTML5 Semantic Tags: Paving the Cow Paths

When HTML5 was being created, the authors faced a tough decision-what tags should they create? There was an obvious desire to clean up HTML as it stood-finally abolishing those tags that had been replaced by equivalent CSS technology. There was also interest in creating tags that replicated some of what had made Flash so successful. Lastly, there was also a desire to create a new round of semantic structural tags. But what should these new structural tags be? One strategy, advocated by Ian Hickson and others was to build structural tags by "paving the cow paths".

What does "paving the cow paths" mean? It refers to a strategy of deciding where to place roads. The idea is that you place roads where there is already a good deal of traffic-you look for the paths that people are already using, and then pave those. This way, you make it easier for people to do what they were already doing, instead of trying to force them go down a different route.
Once the "paving the cow paths" strategy was agreed upon, how was it to be accomplished? Ideally, the authors of the HTML5 specification could look at a database of millions of web pages and build statistics on what popular id and class names were on various tags. Those names would give a good indication of what people were already using. Luckily, Google just happened to have such a database.
When Google ran reports on what people were using, the results were hardly surprising. Web site creators were typically breaking their pages into sections. Pages often had header and footer areas. Many sites (this one included) dealt with publishing information as articles. And almost all sites had some sort of navigation. This report formed the basis of many of the semantic tags in HTML5 today, and we'll go over them in the coming weeks.