>       The first version is correct, since it is a local file. Strapping
> the file:// protocol specifier on it misleads the parser. Remember, this
> is a *PARSER* not a browser, so there's no need to put the file://
> inclusion on it. Unix and Linux are going to need a slightly different
> syntax, but the same logic applies.

But since HTML is written for browsers, and the parser is emulating a
browser, we should probably deal with the browser form as well.

Bill

Reply via email to