A couple of recent articles made me realize the importance of being good enough, but no better.

David Weinberger has a great note on the battle between those who want to make the underlying structure of the web more…structured, and those who see the advantage of the current markup mess.

Dave points out that it is precisely that inattention to well-formed and formal syntax that made the Web a runaway hit with people who were more interested in publishing than in coding. In the very early days of the Web, SGML veterans detested the sloppy HTML we were all writing. The early versions of Netscape were notoriously forgiving of bad HTML and we all benefited:

I mean, if the early browsers only read well-formed and valid HTML, the Web would be far neater, one-thousandth the size, and lifeless.

Meanwhile, Clay Shirky has yet another excellent essay, this time on the tension between “the [vision of the Net] everyone wants — ubiquitous and convenient — and the … the one we get — spotty and cobbled together.”

Clay’s point is that cheap and adequate networks nearly always offer better price/performance than expensive, optimized networks. And these networks can be incrementally optimized over time at a much lower, widely-distributed cost (and faster revenue ramp) than building the perfect network from scratch.

These “nearlynets” have the advantage that they are much less subject to abuse because they’re not owned by a single, inevitably large, investor/builder.