Google Chrome seems to have Problems in rendering a PDF file, if the URL to that file contains non-ascii and non-utf8 url-encoded chars.
This is reproducible in a very simple way:
Chrome will not render the PDF file.
If you add an utf8 string that is url-endcoded it'll work: e.g: '?a=%C3%B6'
This is confirmed with newest Google Chrome and Chromium (45.0) versions on Windows and Linux.
I read this blog, but didn't find a hint that non-utf8 encodings wouldn't work.
If i disable the 'PDF viewer Plugin' (in about:plugins) then the PDF is getting downloaded at least.
Any ideas about this?