>
> So, the same character is treated differently purely because of the way the character is lexically represented.
>
> Isn't that weird?
>
On the contrary, it's exactly what you would expect.
Perhaps in an ideal world there would be symbols that could only be used for markup, completely separate from the symbols used for text.
But then you wouldn't be able to write a markup guide that talks about the symbols used for markup.
As soon as you decide that markup is done using characters that can also be used as ordinary data, you realise that you need escape mechanisms to avoid ambiguities.
Michael Kay
Saxonica
