0.71% of HTML pages valid.

Ben Meadowcroft in this blog entry pointed out a Masters Thesis on parsing invalid HTML which included results saying that 0.71% of the sample URIs were valid. Hopefully people will realise that people can’t author valid documents, so either we need to make our tools ensure validity, or not require validity. Unfortunately the XML world has gone further down the line of requiring validity or the poor user gets nothing, yet I’ve yet to see any attempts at delivering tools that can do the enforcing, people seem to think it’s the UA’s job, it’s NOT, the UA’s job is to show the user the content, not educate the developer.