Could the problem be with FR?
It should work.
Those are curly quotes, e.g.,
“refugees”
The page is supposedly encoded in UTF-8. Curly quotes require three bytes each in UTF-8 (in hex: e2 80 9c and e2 80 9d). A browser rendering UTF-8 should show those triplets as the left and right curly quote characters, respectively.
However, if you look at the page source, what you see is not the above hex, but instead HTML entities corresponding to the three bytes of each triplet:
Muslim “refugees” that are
It appears to be an FR bug. The server is translating individual bytes outside the 7-bit ASCII range to the corresponding HTML entities, resulting in the garbage we are seeing. It should leave them alone and let the browser render them correctly — something browsers have been capable of doing for some years now.
This bug is a recent development. For instance, you now get garbage if you post Russian:
Путин instead of Путин.
Posting Russian (or pretty much anything else) always used to work fine.
