I said that XHTML cleans up the issue, not that it made it simpler. And you are right, XHTML blew their chance by failing to specify a good solution to the problem. However, XHTML at least has a manditory specification standard.

With HTML the browser really has to guess in many cases. The current method of reading the file till you find a content-type tag and then restarting the process of reading the file in the specified type is horribly ugly and depends on no non-ASCII characters being put at the top of the file.

Jay