> i checked the file with an HEX editor and it turned out that 
> there are three
> bytes whose hex code is EF BB BF at the beginning of the 
> offending file.
> I guess this must be the reason why the parser is 
> complaining, though I am
> still not clear if this is some sort of multi-byte character 
> or just some
> junk that happens to be there. 

This is a byte-order-mark, a signal that the file is in UTF-8 encoding. Some
parsers allow a BOM at the start of a UTF-8 file, others don't. The latest
specs say that it should be accepted, but older parsers don't recognize it.
It's generated automatically by some software that writes UTF-8, especially
Microsoft software.

Michael Kay
http://www.saxonica.com/