Hmmm. There is still someting fishy going on with the language files. It seems that the byte in the xml file, to indicate that the content is utf-8 is not particularly liked by Chrome.
If I strip that out, then it seems to work.
I don't know which version of Xerte online toolkits you use, but you can try the following:
Make a backup of modules/xerte/parent_templates/Nottingham/common_html5/js/xenith.js
Edit that file, and search for 'langxmlstr'. You should see a code fragmant like this:
// function gets data from language file
function x_getLangData(lang) {
if (typeof langxmlstr != 'undefined')
{
// We have a off-line object with the language definition in a string
// Convert to an XML object and continue like before
x_languageData = $($.parseXML(langxmlstr)).find("language");
x_setUp();
}
Insert the following line right after the comment that starts with //Convert
langxmlstr = langxmlstr.substr(langxmlstr.indexOf("<"), langxmlstr.lastIndexOf(">") + 1);
So the resulting fragment is as follows:
// function gets data from language file
function x_getLangData(lang) {
if (typeof langxmlstr != 'undefined')
{
// We have a off-line object with the language definition in a string
// Convert to an XML object and continue like before
langxmlstr = langxmlstr.substr(langxmlstr.indexOf("<"), langxmlstr.lastIndexOf(">") + 1);
x_languageData = $($.parseXML(langxmlstr)).find("language");
x_setUp();
}
Of course you can check that in your project, by editing the xenith.js in there first and see if it works for you.
I attached a fixed xenith.js, that I patched from your zip file. You should also be able to copy that in your xerte installation.
