> I believe the HTML 4 spec supports my point That document is badly-written; a more reasonable way to interpret it is to conclude that user agents will use so

> I believe the HTML 4 spec supports my point That document is badly-written; a more reasonable way to interpret it is to conclude that user agents will use some form of Unicode internally, after converting whatever character encoding the document they received used. Which is, indeed, a very reasonable way to design your software, but it doesn't make Latin-1 (for example) a Unicode encoding by any reasonable standard. > A lot of people seem to confuse Unicode with the UTF-encodings. True. I do not.