It’s not just quotes but other characters too that show as question marks. You might have noticed some weeks ago FR had issues with warnings people were getting when posting. Evidently part of the solution was changing the encoding used on the server. It needs to be switched back to UTF-8. I am sure using a third party to have a record of a persons postings paired with their internet address is not an appealing solution to many.
I checked the page. At first glance examination it does not seem to record anything. Even though the key part of the code is very short:
s = s.replace( /\u2018|\u2019|\u201A|\uFFFD/g, “’” );
s = s.replace( /\u201c|\u201d|\u201e/g, ‘”’ );
s = s.replace( /\u02C6/g, ‘^’ );
s = s.replace( /\u2039/g, ‘<’ );
s = s.replace( /\u203A/g, ‘>’ );
s = s.replace( /\u2013/g, ‘-’ );
s = s.replace( /\u2014/g, ‘—’ );
s = s.replace( /\u2026/g, ‘...’ );
s = s.replace( /\u00A9/g, ‘(c)’ );
s = s.replace( /\u00AE/g, ‘(r)’ );
s = s.replace( /\u2122/g, ‘TM’ );
s = s.replace( /\u00BC/g, ‘1/4’ );
s = s.replace( /\u00BD/g, ‘1/2’ );
s = s.replace( /\u00BE/g, ‘3/4’ );
s = s.replace(/[\u02DC|\u00A0]/g, “ “);
FR should be able to easily incorporate it into their code.