>Client-side javascript in my app does a lot of XMLHTTP calls to collect snippets of HTML and refresh portions of the web page using the ResponseText property of the xmlhttp object.
>
>Currently, the web server simply spits out HTML snippets in response to either a post or get request from the xmlhttp object. This works fine. However, there is a known IE bug when non-ascii characters are transmitted in the ResponseText. Those characters get garbled.
>
>I've read that you can use XML (using the ResponseXML property of the xmlhttp object) to circumvent the non-ascii-character bug by encapsulting the HTML snippet within an XML package. I just need some pointers on how to actually do this... especially what the server needs to do to create acceptable XML with the appropriate HTTP headers that all gets sent back to the xmlhttp object for parsing on the client.
>
>All I'm trying to do is get around the problem where certain European/Latin characters (high order bit set in single-byte character) get munged when using ResponseText.
>
>Thanks very much!

In continuing from my last reply...

Also try "application/xml". The "applcation" portion tells the server that the response may be binary (8 bits), where the "text" type is only 7-bits.