We have a website (www.semanticdesigns.com) that delivers a default encoding of ISO-8859-1. The pages are written in, guess what, ASCII mostly with some ISO-8859-1 characters.
We have a few example directories where the pages under that directory are actually UTF-8-encoded with BOM. (http://www.semanticdesigns.com/Products/Formatters/JavaHTMLJunit/index.html) These display fine in IE8 and IE9, but FireFox gets confused and displays just the BOM characters
" "
My reading of the HTML standard is that we need to add a meta tag to those pages:
<meta http-equiv="content-type" content="text/html;charset=UTF-8"></meta>
These pages are generated automatically if it isn't obvious, so this is an easy change to make. But, I tried patching this by hand, but it doesn't seem to have an effect on FireFox.
How does the browser actually decide which encoding to use? The default one provided by the server? Or what's in the meta tag? Suggestions on how to solve this problem, without changing the site default?