[Coda]

Computers can only deal in 1's and 0's. In both Latin-1 and UTF-8 (two different ways of storing text), an uppercase letter A is stored as 01000001 (that is, 65). In UTF-8, ð is stored as 1100001110110000. However, if you read those 1's and 0's using Latin-1, 11000011 is à and 10110000 is °.

Where things REALLY start getting confused is if you have some part of the software that thinks the data is stored on-disk in Latin-1 and wants to translate it into UTF-8 to talk to your web browser. Then it takes what it thinks is "ð" and converts it into how UTF-8 would express those... 1100001110000011 for à and 1100001010110000 for °. Because that's what it would have to send to a web browser if you WANTED to see "ð".