User-Agent: Mozilla/5.0 (Macintosh; U; PPC Mac OS X Mach-O; en-US; rv:1.7.5) Gecko/20041107 Firefox/1.0 Build Identifier: Mozilla/5.0 (Macintosh; U; PPC Mac OS X Mach-O; en-US; rv:1.7.5) Gecko/20041107 Firefox/1.0 When using charset parameter using quoted-string, Mozilla/Firefox ignore the charset and use the charset defined in the head meta tag. Example: Content-Type: text/html; charset="iso-8859-8" According to href="http://www.w3.org/Protocols/rfc2616/rfc2616-sec3.html#sec3.7, media type e.g. text/html can use parameters using format specified in http://www.w3.org/Protocols/rfc2616/rfc2616-sec3.html#sec3.6. The quoting is not needed in this case, but this seems to be valid HTTP 1.1 content type declaration. * IE 6.0 also ignore such charset * Safari respect it * W3C HTML validator respect this charset Compare with same testcase using token: 'Content-Type: text/html; charset=utf-8'. Here Mozilla works fine. http://nirs.dyndns.org/cgi-bin/token.cgi Reproducible: Always Steps to Reproduce: 1. Visit testcase url Actual Results: Page rendered using charset defined in head meta tag Expected Results: Page should render in charset defined in HTTP header Content-type

The source of my Python cgi testcase.

*** This bug has been marked as a duplicate of 244964 ***

V/dupe