lucy24 - 9:56 pm on Jul 28, 2011 (gmt 0)

It's a horrible solution, but if you express all non-ASCII chracters as html entities a word my fingers flatly refuse to type right on the first try then the stated file encoding doesn't matter.

This I did, however when I write cañon guía, in the email I get cañon guía.

Yup. What you've got there is UTF-8 encoded text being reinterpreted as ISO-Latin-1.

What language is the database in? Not what human language ;) I mean what computer language? There's almost always some type of encode/decode function to deal with non-ASCII characters, or characters that need to be escaped for other reasons.

And remember that changing the html header's "charset" does not change the actual file, it only changes how the browser reads it. So it has to be composed in-- or changed to-- UTF-8 in the first place.

:: happily patting SubEthaEdit on the head because it will both convert and reinterpret on the fly ::