[Openmcl-devel] how many angels can dance on a unicode character?

[Hamilton Link]

On Apr 21, 2007, at 3:40 AM, Gary Byers wrote:

> ... OpenMCL ("best thing since a clay tablet!") ...
>

This is a good example of why I read openmcl-devel so religiously.

But the argument about really, really, REALLY wanting to keep 
destructive string operations simple and fast is a good one.  Without 
just saying surrogate code points are "characters" and allowing strings 
of first-halves-of-cuneform-characters and other such nonsense I guess 
UTF32 is moderately more sane (I assume this is the unpleasantness Gary 
didn't want to consider... I don't know what CLTL2 says about "what it 
means to be a character" or what UTF says about "what it means to be a 
character" so this unpleasantness may be in violation of multiple 
formal definitions).  Oh, and where can I get a Cuneiform font for Mac 
OS X, anyone know?  Because this eMac sure beats the heck out of a clay 
tablet.

On a side note (in regards to people using odd non-UCS2 characters), C# 
is called "C#" rather than "C<unicode musical sharp whatever that is>" 
even though .net uses UTF16 (and deals almost exclusively with 
immutable strings, by the way), on the grounds that for most people # 
is easier to type than the unicode musical sharp.

h