For compatibility with the original Squeak, ideally the definition of the one-octet characters and strings should not be changed. However, to adapt to Unicode, it is cleaner if the first 256 characters are identical with ISO-8859-1 character set, instead of the original Squeak's MacRoman character set.
We decided to modify the upper half of the first 256 characters to be ISO-8859-1 characters. Because the characters in the upper half are not used much, this change was easy. With this change, the Character class represents the ISO-8859-1, which is equivalent to the first 256 characters in Unicode standard.