Character encoding (UTF-8 vs ISO-8859-1)

Hello, The default encoding for the PHP manual is ISO-8859-1. Some other languages get special encodings (you can find these in configure.in, around line 655). This gives problems for some common languages. If you take a look at the French translation of the echo function [ http://php.net/manual/fr/function.echo.php ], you see the character entities (like &eacute;) in the code example aren't replaced by the correct character. This is because in the XML source file, code examples appear inside <![CDATA[ ]]> sections, and entities aren't read there. If we could place the correct character instead of the entity there (and in the rest of the manual), these problems wouldn't occur. I have noticed that this gives problems if the compilation happens in the default ISO-8859-1 encoding, but not with UTF-8. What would be the implication of changing the default encoding from ISO-8859-1 to UTF-8? What would break? What changes in the source files would be needed? I have searched the list archives for an earlier discussion of this subject, but couldn't find anything. Greetings, Jan Fabry