Problem with url-retrieve-synchronously and HTTP header

["Christian Henz" <ChrHenz@gmx.net>]

In emacs 23.4.1, I'm trying to access a local web service (written in Python/Flask), but "url-retrieve-synchronously" does not properly process the HTTP header returned.

When I retrieve google.com (among other sites), it works as expected:

    (switch-to-buffer (url-retrieve-synchronously "http://google.com"))
    HTTP/1.1 200 OK
    Date: Thu, 06 Feb 2014 22:15:10 GMT
    [...]

When I retrieve my local URL, the header is not parsed/cleaned correctly:

    (switch-to-buffer (url-retrieve-synchronously "http://127.0.0.1:5000"))
    HTTP/1.0 200 OK
    Content-Type: text/plain; charset=UTF-8^M
    [...]

Only the first line was "cleaned" by converting "\r\n" -> "\n". "url-http-end-of-headers" then consequently points to the "\n" after "OK".

When I dump the header using curl, it looks correct to me:

    HTTP/1.0 200 OK\r\n
    Content-Type: text/plain; charset=UTF-8\r\n
    [...]

I also remember this working as expected on another machine with the same local service, but emacs 22...

Does anyone know what the problem might be?

Cheers,
Christian