Jonathan S. Shapiro wrote:
> On Sun, 2006-05-21 at 17:54 +0200, Marcin 'Qrczak' Kowalczyk wrote:
>> "Jonathan S. Shapiro" <xxxxxx@eros-os.org> writes:
>>
>>>   1. The correct primitive is READ-CODEPOINT, not READ-CHAR.
>>>      READ-CHAR is a library routine.
>> Unless char is defined to be a code point. Which is IMHO the most
>> reasonable choice: code points are the natural atomic units of Unicode
>> text, and most Unicode algorithms are expressed in terms of code points.
>
> In many respects I agree that this would be sensible from the
> programmer's perspective.
>
> Unfortunately it is quite wrong, which is something that the UNICODE
> people go to great lengths to make clear (and, IMO, a serious failing of
> UNICODE).

I don't think that is relevant.  The point is that the Scheme concept of
character is most sensibly mapped to the Unicode concept of codepoint.
We should stay far away from attempting any kind of data-type that tries
models character combinations - for that people should use strings.

Hence, there is no need to use the name "read-codepoint" - "read-char"
is just fine, since it reads a Scheme character, which happens to be
the same as a Unicode codepoint.
--
	--Per Bothner
xxxxxx@bothner.com   http://per.bothner.com/